diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index d0e070e15..e042683de 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -15,6 +15,7 @@ env:
   DEFAULT_UV_VERSION: "0.8.x"
   # This is the default version of Python to use in most steps which aren't specific
   DEFAULT_PYTHON_VERSION: "3.11"
+  NLTK_DATA: "/usr/share/nltk_data"
 jobs:
   pre-commit:
     # We want to run on external PRs, but not on our own internal PRs as they'll be run
@@ -121,8 +122,11 @@ jobs:
       - name: List installed Python dependencies
         run: |
           uv pip list
+      - name: Install or update NLTK dependencies
+        run: uv run python -m nltk.downloader punkt punkt_tab snowball_data stopwords -d ${{ env.NLTK_DATA }}
       - name: Tests
         env:
+          NLTK_DATA: ${{ env.NLTK_DATA }}
           PAPERLESS_CI_TEST: 1
           # Enable paperless_mail testing against real server
           PAPERLESS_MAIL_TEST_HOST: ${{ secrets.TEST_MAIL_HOST }}
diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index 64a183186..2c50e81a6 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -31,7 +31,7 @@ repos:
     rev: v2.4.1
     hooks:
       - id: codespell
-        exclude: "(^src-ui/src/locale/)|(^src-ui/pnpm-lock.yaml)|(^src-ui/e2e/)|(^src/paperless_mail/tests/samples/)"
+        exclude: "(^src-ui/src/locale/)|(^src-ui/pnpm-lock.yaml)|(^src-ui/e2e/)|(^src/paperless_mail/tests/samples/)|(^src/documents/tests/samples/)"
         exclude_types:
           - pofile
           - json
diff --git a/docs/configuration.md b/docs/configuration.md
index 92500d26c..9b7041962 100644
--- a/docs/configuration.md
+++ b/docs/configuration.md
@@ -159,6 +159,23 @@ Available options are `postgresql` and `mariadb`.
 
     Defaults to unset, which uses Django’s built-in defaults.
 
+#### [`PAPERLESS_DB_POOLSIZE=<int>`](#PAPERLESS_DB_POOLSIZE) {#PAPERLESS_DB_POOLSIZE}
+
+: Defines the maximum number of database connections to keep in the pool.
+
+    Only applies to PostgreSQL. This setting is ignored for other database engines.
+
+    The value must be greater than or equal to 1 to be used.
+    Defaults to unset, which disables connection pooling.
+
+    !!! note
+
+    A small pool is typically sufficient — for example, a size of 4.
+    Make sure your PostgreSQL server's max_connections setting is large enough to handle:
+    ```(Paperless workers + Celery workers) × pool size + safety margin```
+    For example, with 4 Paperless workers and 2 Celery workers, and a pool size of 4:
+    (4 + 2) × 4 + 10 = 34 connections required.
+
 #### [`PAPERLESS_DB_READ_CACHE_ENABLED=<bool>`](#PAPERLESS_DB_READ_CACHE_ENABLED) {#PAPERLESS_DB_READ_CACHE_ENABLED}
 
 : Caches the database read query results into Redis. This can significantly improve application response times by caching database queries, at the cost of slightly increased memory usage.
diff --git a/docs/usage.md b/docs/usage.md
index 5c377a466..a2686ab7d 100644
--- a/docs/usage.md
+++ b/docs/usage.md
@@ -30,6 +30,9 @@ Each document has data fields that you can assign to them:
 -   A _document type_ is used to demarcate the type of a document such
     as letter, bank statement, invoice, contract, etc. It is used to
     identify what a document is about.
+-   The document _storage path_ is the location where the document files
+    are stored. See [Storage Paths](advanced_usage.md#storage-paths) for
+    more information.
 -   The _date added_ of a document is the date the document was scanned
     into paperless. You cannot and should not change this date.
 -   The _date created_ of a document is the date the document was
diff --git a/pyproject.toml b/pyproject.toml
index 26f192231..4b7aaa68d 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -53,6 +53,7 @@ dependencies = [
   "ocrmypdf~=16.10.0",
   "pathvalidate~=3.3.1",
   "pdf2image~=1.17.0",
+  "psycopg-pool",
   "python-dateutil~=2.9.0",
   "python-dotenv~=1.1.0",
   "python-gnupg~=0.5.4",
@@ -63,7 +64,7 @@ dependencies = [
   "redis[hiredis]~=5.2.1",
   "scikit-learn~=1.7.0",
   "setproctitle~=1.3.4",
-  "tika-client~=0.9.0",
+  "tika-client~=0.10.0",
   "tqdm~=4.67.1",
   "watchdog~=6.0",
   "whitenoise~=6.9",
@@ -75,9 +76,10 @@ optional-dependencies.mariadb = [
   "mysqlclient~=2.2.7",
 ]
 optional-dependencies.postgres = [
-  "psycopg[c]==3.2.9",
+  "psycopg[c,pool]==3.2.9",
   # Direct dependency for proper resolution of the pre-built wheels
   "psycopg-c==3.2.9",
+  "psycopg-pool==3.2.6",
 ]
 optional-dependencies.webserver = [
   "granian[uvloop]~=2.4.1",
@@ -203,15 +205,9 @@ lint.per-file-ignores."docker/wait-for-redis.py" = [
   "INP001",
   "T201",
 ]
-lint.per-file-ignores."src/documents/file_handling.py" = [
-  "PTH",
-] # TODO Enable & remove
 lint.per-file-ignores."src/documents/management/commands/document_consumer.py" = [
   "PTH",
 ] # TODO Enable & remove
-lint.per-file-ignores."src/documents/management/commands/document_exporter.py" = [
-  "PTH",
-] # TODO Enable & remove
 lint.per-file-ignores."src/documents/migrations/1012_fix_archive_files.py" = [
   "PTH",
 ] # TODO Enable & remove
@@ -221,9 +217,6 @@ lint.per-file-ignores."src/documents/models.py" = [
 lint.per-file-ignores."src/documents/parsers.py" = [
   "PTH",
 ] # TODO Enable & remove
-lint.per-file-ignores."src/documents/signals/handlers.py" = [
-  "PTH",
-] # TODO Enable & remove
 lint.per-file-ignores."src/paperless_tesseract/tests/test_parser.py" = [
   "RUF001",
 ]
diff --git a/src-ui/messages.xlf b/src-ui/messages.xlf
index 7b26a3e7a..5bf7c66e1 100644
--- a/src-ui/messages.xlf
+++ b/src-ui/messages.xlf
@@ -332,19 +332,19 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">102</context>
+          <context context-type="linenumber">103</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">102</context>
+          <context context-type="linenumber">103</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">102</context>
+          <context context-type="linenumber">103</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">102</context>
+          <context context-type="linenumber">103</context>
         </context-group>
       </trans-unit>
       <trans-unit id="4930506384627295710" datatype="html">
@@ -545,7 +545,7 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/settings/settings.component.html</context>
-          <context context-type="linenumber">361</context>
+          <context context-type="linenumber">362</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/common/edit-dialog/correspondent-edit-dialog/correspondent-edit-dialog.component.html</context>
@@ -605,7 +605,7 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/saved-views/saved-views.component.html</context>
-          <context context-type="linenumber">73</context>
+          <context context-type="linenumber">74</context>
         </context-group>
       </trans-unit>
       <trans-unit id="5079885666748292382" datatype="html">
@@ -763,19 +763,19 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">51</context>
+          <context context-type="linenumber">52</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">51</context>
+          <context context-type="linenumber">52</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">51</context>
+          <context context-type="linenumber">52</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">51</context>
+          <context context-type="linenumber">52</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/saved-views/saved-views.component.html</context>
@@ -1225,19 +1225,19 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">6</context>
+          <context context-type="linenumber">7</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">6</context>
+          <context context-type="linenumber">7</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">6</context>
+          <context context-type="linenumber">7</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">6</context>
+          <context context-type="linenumber">7</context>
         </context-group>
       </trans-unit>
       <trans-unit id="309314153079578337" datatype="html">
@@ -1432,7 +1432,7 @@
         <source>Cancel</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/admin/settings/settings.component.html</context>
-          <context context-type="linenumber">362</context>
+          <context context-type="linenumber">361</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/common/confirm-dialog/confirm-dialog.component.ts</context>
@@ -1500,7 +1500,7 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/saved-views/saved-views.component.html</context>
-          <context context-type="linenumber">74</context>
+          <context context-type="linenumber">73</context>
         </context-group>
       </trans-unit>
       <trans-unit id="6839066544204061364" datatype="html">
@@ -1598,19 +1598,19 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">3</context>
+          <context context-type="linenumber">4</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">3</context>
+          <context context-type="linenumber">4</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">3</context>
+          <context context-type="linenumber">4</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">3</context>
+          <context context-type="linenumber">4</context>
         </context-group>
       </trans-unit>
       <trans-unit id="4880728824338713664" datatype="html">
@@ -1696,35 +1696,35 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">20</context>
+          <context context-type="linenumber">21</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">20</context>
+          <context context-type="linenumber">21</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">20</context>
+          <context context-type="linenumber">21</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">20</context>
+          <context context-type="linenumber">21</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">37</context>
+          <context context-type="linenumber">38</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">37</context>
+          <context context-type="linenumber">38</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">37</context>
+          <context context-type="linenumber">38</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">37</context>
+          <context context-type="linenumber">38</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/workflows/workflows.component.html</context>
@@ -1816,19 +1816,19 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">43</context>
+          <context context-type="linenumber">44</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">43</context>
+          <context context-type="linenumber">44</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">43</context>
+          <context context-type="linenumber">44</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">43</context>
+          <context context-type="linenumber">44</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/saved-views/saved-views.component.html</context>
@@ -2121,51 +2121,51 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">9</context>
+          <context context-type="linenumber">10</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">9</context>
+          <context context-type="linenumber">10</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">9</context>
+          <context context-type="linenumber">10</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">9</context>
+          <context context-type="linenumber">10</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">84</context>
+          <context context-type="linenumber">85</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">84</context>
+          <context context-type="linenumber">85</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">84</context>
+          <context context-type="linenumber">85</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">84</context>
+          <context context-type="linenumber">85</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">96</context>
+          <context context-type="linenumber">97</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">96</context>
+          <context context-type="linenumber">97</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">96</context>
+          <context context-type="linenumber">97</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">96</context>
+          <context context-type="linenumber">97</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.ts</context>
@@ -2440,35 +2440,35 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">83</context>
+          <context context-type="linenumber">84</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">83</context>
+          <context context-type="linenumber">84</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">83</context>
+          <context context-type="linenumber">84</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">83</context>
+          <context context-type="linenumber">84</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">93</context>
+          <context context-type="linenumber">94</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">93</context>
+          <context context-type="linenumber">94</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">93</context>
+          <context context-type="linenumber">94</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">93</context>
+          <context context-type="linenumber">94</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/workflows/workflows.component.html</context>
@@ -5227,19 +5227,19 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">12</context>
+          <context context-type="linenumber">13</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">12</context>
+          <context context-type="linenumber">13</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">12</context>
+          <context context-type="linenumber">13</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">12</context>
+          <context context-type="linenumber">13</context>
         </context-group>
       </trans-unit>
       <trans-unit id="4391289919356861627" datatype="html">
@@ -8333,19 +8333,19 @@
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">86</context>
+          <context context-type="linenumber">87</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">86</context>
+          <context context-type="linenumber">87</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">86</context>
+          <context context-type="linenumber">87</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">86</context>
+          <context context-type="linenumber">87</context>
         </context-group>
       </trans-unit>
       <trans-unit id="651372623796033489" datatype="html">
@@ -8672,76 +8672,76 @@
         <source>Filter by:</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">19</context>
+          <context context-type="linenumber">20</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">19</context>
+          <context context-type="linenumber">20</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">19</context>
+          <context context-type="linenumber">20</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">19</context>
+          <context context-type="linenumber">20</context>
         </context-group>
       </trans-unit>
       <trans-unit id="1383365546483928780" datatype="html">
         <source>Matching</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">38</context>
+          <context context-type="linenumber">39</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">38</context>
+          <context context-type="linenumber">39</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">38</context>
+          <context context-type="linenumber">39</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">38</context>
+          <context context-type="linenumber">39</context>
         </context-group>
       </trans-unit>
       <trans-unit id="1488347670280290838" datatype="html">
         <source>Document count</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">39</context>
+          <context context-type="linenumber">40</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">39</context>
+          <context context-type="linenumber">40</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">39</context>
+          <context context-type="linenumber">40</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">39</context>
+          <context context-type="linenumber">40</context>
         </context-group>
       </trans-unit>
       <trans-unit id="8095412801504464756" datatype="html">
         <source>{VAR_PLURAL, plural, =1 {One <x id="INTERPOLATION"/>} other {<x id="INTERPOLATION_1"/> total <x id="INTERPOLATION_2"/>}}</source>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">118</context>
+          <context context-type="linenumber">119</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">118</context>
+          <context context-type="linenumber">119</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">118</context>
+          <context context-type="linenumber">119</context>
         </context-group>
         <context-group purpose="location">
           <context context-type="sourcefile">src/app/components/manage/management-list/management-list.component.html</context>
-          <context context-type="linenumber">118</context>
+          <context context-type="linenumber">119</context>
         </context-group>
       </trans-unit>
       <trans-unit id="810888510148304696" datatype="html">
diff --git a/src-ui/src/app/components/admin/config/config.component.html b/src-ui/src/app/components/admin/config/config.component.html
index 0f74339fb..e1d7340a6 100644
--- a/src-ui/src/app/components/admin/config/config.component.html
+++ b/src-ui/src/app/components/admin/config/config.component.html
@@ -50,7 +50,7 @@
     <div [ngbNavOutlet]="nav" class="border-start border-end border-bottom p-3 mb-3 shadow-sm"></div>
     <div class="btn-toolbar" role="toolbar">
         <div class="btn-group me-2">
-            <button type="button" (click)="discardChanges()" class="btn btn-secondary" [disabled]="loading || (isDirty$ | async) === false" i18n>Discard</button>
+            <button type="button" (click)="discardChanges()" class="btn btn-outline-secondary" [disabled]="loading || (isDirty$ | async) === false" i18n>Discard</button>
         </div>
         <div class="btn-group">
             <button type="submit" class="btn btn-primary" [disabled]="loading || !configForm.valid || (isDirty$ | async) === false" i18n>Save</button>
diff --git a/src-ui/src/app/components/admin/settings/settings.component.html b/src-ui/src/app/components/admin/settings/settings.component.html
index ccd3cc7e3..b228dac32 100644
--- a/src-ui/src/app/components/admin/settings/settings.component.html
+++ b/src-ui/src/app/components/admin/settings/settings.component.html
@@ -358,6 +358,6 @@
 
   <div [ngbNavOutlet]="nav" class="border-start border-end border-bottom p-3 mb-3 shadow-sm"></div>
 
-  <button type="submit" class="btn btn-primary mb-2" [disabled]="(isDirty$ | async) === false" i18n>Save</button>
-  <button type="button" (click)="reset()" class="btn btn-secondary ms-2 mb-2" [disabled]="(isDirty$ | async) === false" i18n>Cancel</button>
+  <button type="button" (click)="reset()" class="btn btn-outline-secondary mb-2" [disabled]="(isDirty$ | async) === false" i18n>Cancel</button>
+  <button type="submit" class="btn btn-primary ms-2 mb-2" [disabled]="(isDirty$ | async) === false" i18n>Save</button>
 </form>
diff --git a/src-ui/src/app/components/manage/management-list/management-list.component.html b/src-ui/src/app/components/manage/management-list/management-list.component.html
index 82fd8502c..6375a3667 100644
--- a/src-ui/src/app/components/manage/management-list/management-list.component.html
+++ b/src-ui/src/app/components/manage/management-list/management-list.component.html
@@ -1,4 +1,5 @@
-<pngx-page-header title="{{ typeNamePlural | titlecase }}">
+<pngx-page-header title="{{ typeNamePlural | titlecase }}" info="View, add, edit and delete {{ typeNamePlural }}." infoLink="usage/#terms-and-definitions">
+
   <button class="btn btn-sm btn-outline-secondary" (click)="clearSelection()" [hidden]="selectedObjects.size === 0">
     <i-bs  name="x"></i-bs>&nbsp;<ng-container i18n>Clear selection</ng-container>
     </button>
diff --git a/src-ui/src/app/components/manage/management-list/management-list.component.spec.ts b/src-ui/src/app/components/manage/management-list/management-list.component.spec.ts
index 291a71fa3..7df6cd5a5 100644
--- a/src-ui/src/app/components/manage/management-list/management-list.component.spec.ts
+++ b/src-ui/src/app/components/manage/management-list/management-list.component.spec.ts
@@ -164,7 +164,7 @@ describe('ManagementListComponent', () => {
     const toastInfoSpy = jest.spyOn(toastService, 'showInfo')
     const reloadSpy = jest.spyOn(component, 'reloadData')
 
-    const createButton = fixture.debugElement.queryAll(By.css('button'))[3]
+    const createButton = fixture.debugElement.queryAll(By.css('button'))[4]
     createButton.triggerEventHandler('click')
 
     expect(modal).not.toBeUndefined()
@@ -188,7 +188,7 @@ describe('ManagementListComponent', () => {
     const toastInfoSpy = jest.spyOn(toastService, 'showInfo')
     const reloadSpy = jest.spyOn(component, 'reloadData')
 
-    const editButton = fixture.debugElement.queryAll(By.css('button'))[6]
+    const editButton = fixture.debugElement.queryAll(By.css('button'))[7]
     editButton.triggerEventHandler('click')
 
     expect(modal).not.toBeUndefined()
@@ -213,7 +213,7 @@ describe('ManagementListComponent', () => {
     const deleteSpy = jest.spyOn(tagService, 'delete')
     const reloadSpy = jest.spyOn(component, 'reloadData')
 
-    const deleteButton = fixture.debugElement.queryAll(By.css('button'))[7]
+    const deleteButton = fixture.debugElement.queryAll(By.css('button'))[8]
     deleteButton.triggerEventHandler('click')
 
     expect(modal).not.toBeUndefined()
@@ -233,7 +233,7 @@ describe('ManagementListComponent', () => {
 
   it('should support quick filter for objects', () => {
     const qfSpy = jest.spyOn(documentListViewService, 'quickFilter')
-    const filterButton = fixture.debugElement.queryAll(By.css('button'))[8]
+    const filterButton = fixture.debugElement.queryAll(By.css('button'))[9]
     filterButton.triggerEventHandler('click')
     expect(qfSpy).toHaveBeenCalledWith([
       { rule_type: FILTER_HAS_TAGS_ALL, value: tags[0].id.toString() },
diff --git a/src-ui/src/app/components/manage/saved-views/saved-views.component.html b/src-ui/src/app/components/manage/saved-views/saved-views.component.html
index 296cf71ac..10487fec8 100644
--- a/src-ui/src/app/components/manage/saved-views/saved-views.component.html
+++ b/src-ui/src/app/components/manage/saved-views/saved-views.component.html
@@ -70,6 +70,6 @@
     }
   </ul>
 
-  <button type="submit" class="btn btn-primary mb-2" [disabled]="(isDirty$ | async) === false" i18n>Save</button>
-  <button type="button" (click)="reset()" class="btn btn-secondary ms-2 mb-2" [disabled]="(isDirty$ | async) === false" i18n>Cancel</button>
+  <button type="button" (click)="reset()" class="btn btn-outline-secondary mb-2" [disabled]="(isDirty$ | async) === false" i18n>Cancel</button>
+  <button type="submit" class="btn btn-primary ms-2 mb-2" [disabled]="(isDirty$ | async) === false" i18n>Save</button>
 </form>
diff --git a/src/documents/caching.py b/src/documents/caching.py
index 1099a7a73..ed7f6dbc1 100644
--- a/src/documents/caching.py
+++ b/src/documents/caching.py
@@ -1,16 +1,23 @@
 from __future__ import annotations
 
 import logging
+import pickle
 from binascii import hexlify
+from collections import OrderedDict
 from dataclasses import dataclass
 from typing import TYPE_CHECKING
+from typing import Any
 from typing import Final
 
+from django.conf import settings
 from django.core.cache import cache
+from django.core.cache import caches
 
 from documents.models import Document
 
 if TYPE_CHECKING:
+    from django.core.cache.backends.base import BaseCache
+
     from documents.classifier import DocumentClassifier
 
 logger = logging.getLogger("paperless.caching")
@@ -39,6 +46,80 @@ CACHE_1_MINUTE: Final[int] = 60
 CACHE_5_MINUTES: Final[int] = 5 * CACHE_1_MINUTE
 CACHE_50_MINUTES: Final[int] = 50 * CACHE_1_MINUTE
 
+read_cache = caches["read-cache"]
+
+
+class LRUCache:
+    def __init__(self, capacity: int = 128):
+        self._data = OrderedDict()
+        self.capacity = capacity
+
+    def get(self, key, default=None) -> Any | None:
+        if key in self._data:
+            self._data.move_to_end(key)
+            return self._data[key]
+        return default
+
+    def set(self, key, value) -> None:
+        self._data[key] = value
+        self._data.move_to_end(key)
+        while len(self._data) > self.capacity:
+            self._data.popitem(last=False)
+
+
+class StoredLRUCache(LRUCache):
+    """
+    LRU cache that can persist its entire contents as a single entry in a backend cache.
+
+    Useful for sharing a cache across multiple workers or processes.
+
+    Workflow:
+        1. Load the cache state from the backend using `load()`.
+        2. Use `get()` and `set()` locally as usual.
+        3. Persist changes back to the backend using `save()`.
+    """
+
+    def __init__(
+        self,
+        backend_key: str,
+        capacity: int = 128,
+        backend: BaseCache = read_cache,
+        backend_ttl=settings.CACHALOT_TIMEOUT,
+    ):
+        if backend_key is None:
+            raise ValueError("backend_key is mandatory")
+        super().__init__(capacity)
+        self._backend_key = backend_key
+        self._backend = backend
+        self.backend_ttl = backend_ttl
+
+    def load(self) -> None:
+        """
+        Load the whole cache content from backend storage.
+
+        If no valid cached data exists in the backend, the local cache is cleared.
+        """
+        serialized_data = self._backend.get(self._backend_key)
+        try:
+            self._data = (
+                pickle.loads(serialized_data) if serialized_data else OrderedDict()
+            )
+        except pickle.PickleError:
+            logger.warning(
+                "Cache exists in backend but could not be read (possibly invalid format)",
+            )
+
+    def save(self) -> None:
+        """Save the entire local cache to the backend as a serialized object.
+
+        The backend entry will expire after the configured TTL.
+        """
+        self._backend.set(
+            self._backend_key,
+            pickle.dumps(self._data),
+            self.backend_ttl,
+        )
+
 
 def get_suggestion_cache_key(document_id: int) -> str:
     """
diff --git a/src/documents/classifier.py b/src/documents/classifier.py
index 728c83228..613c1d5ad 100644
--- a/src/documents/classifier.py
+++ b/src/documents/classifier.py
@@ -16,16 +16,29 @@ if TYPE_CHECKING:
 
 from django.conf import settings
 from django.core.cache import cache
+from django.core.cache import caches
 
+from documents.caching import CACHE_5_MINUTES
 from documents.caching import CACHE_50_MINUTES
 from documents.caching import CLASSIFIER_HASH_KEY
 from documents.caching import CLASSIFIER_MODIFIED_KEY
 from documents.caching import CLASSIFIER_VERSION_KEY
+from documents.caching import StoredLRUCache
 from documents.models import Document
 from documents.models import MatchingModel
 
 logger = logging.getLogger("paperless.classifier")
 
+ADVANCED_TEXT_PROCESSING_ENABLED = (
+    settings.NLTK_LANGUAGE is not None and settings.NLTK_ENABLED
+)
+
+read_cache = caches["read-cache"]
+
+
+RE_DIGIT = re.compile(r"\d")
+RE_WORD = re.compile(r"\b[\w]+\b")  # words that may contain digits
+
 
 class IncompatibleClassifierVersionError(Exception):
     def __init__(self, message: str, *args: object) -> None:
@@ -92,15 +105,28 @@ class DocumentClassifier:
         self.last_auto_type_hash: bytes | None = None
 
         self.data_vectorizer = None
+        self.data_vectorizer_hash = None
         self.tags_binarizer = None
         self.tags_classifier = None
         self.correspondent_classifier = None
         self.document_type_classifier = None
         self.storage_path_classifier = None
-
         self._stemmer = None
+        # 10,000 elements roughly use 200 to 500 KB per worker,
+        # and also in the shared Redis cache,
+        # Keep this cache small to minimize lookup and I/O latency.
+        if ADVANCED_TEXT_PROCESSING_ENABLED:
+            self._stem_cache = StoredLRUCache(
+                f"stem_cache_v{self.FORMAT_VERSION}",
+                capacity=10000,
+            )
         self._stop_words = None
 
+    def _update_data_vectorizer_hash(self):
+        self.data_vectorizer_hash = sha256(
+            pickle.dumps(self.data_vectorizer),
+        ).hexdigest()
+
     def load(self) -> None:
         from sklearn.exceptions import InconsistentVersionWarning
 
@@ -119,6 +145,7 @@ class DocumentClassifier:
                         self.last_auto_type_hash = pickle.load(f)
 
                         self.data_vectorizer = pickle.load(f)
+                        self._update_data_vectorizer_hash()
                         self.tags_binarizer = pickle.load(f)
 
                         self.tags_classifier = pickle.load(f)
@@ -269,7 +296,7 @@ class DocumentClassifier:
             Generates the content for documents, but once at a time
             """
             for doc in docs_queryset:
-                yield self.preprocess_content(doc.content)
+                yield self.preprocess_content(doc.content, shared_cache=False)
 
         self.data_vectorizer = CountVectorizer(
             analyzer="word",
@@ -347,6 +374,7 @@ class DocumentClassifier:
 
         self.last_doc_change_time = latest_doc_change
         self.last_auto_type_hash = hasher.digest()
+        self._update_data_vectorizer_hash()
 
         # Set the classifier information into the cache
         # Caching for 50 minutes, so slightly less than the normal retrain time
@@ -356,30 +384,15 @@ class DocumentClassifier:
 
         return True
 
-    def preprocess_content(self, content: str) -> str:  # pragma: no cover
-        """
-        Process to contents of a document, distilling it down into
-        words which are meaningful to the content
-        """
-
-        # Lower case the document
-        content = content.lower().strip()
-        # Reduce spaces
-        content = re.sub(r"\s+", " ", content)
-        # Get only the letters
-        content = re.sub(r"[^\w\s]", " ", content)
-
-        # If the NLTK language is supported, do further processing
-        if settings.NLTK_LANGUAGE is not None and settings.NLTK_ENABLED:
+    def _init_advanced_text_processing(self):
+        if self._stop_words is None or self._stemmer is None:
             import nltk
             from nltk.corpus import stopwords
             from nltk.stem import SnowballStemmer
-            from nltk.tokenize import word_tokenize
 
             # Not really hacky, since it isn't private and is documented, but
             # set the search path for NLTK data to the single location it should be in
             nltk.data.path = [settings.NLTK_DIR]
-
             try:
                 # Preload the corpus early, to force the lazy loader to transform
                 stopwords.ensure_loaded()
@@ -387,41 +400,100 @@ class DocumentClassifier:
                 # Do some one time setup
                 # Sometimes, somehow, there's multiple threads loading the corpus
                 # and it's not thread safe, raising an AttributeError
-                if self._stemmer is None:
-                    self._stemmer = SnowballStemmer(settings.NLTK_LANGUAGE)
-                if self._stop_words is None:
-                    self._stop_words = set(stopwords.words(settings.NLTK_LANGUAGE))
-
-                # Tokenize
-                # This splits the content into tokens, roughly words
-                words: list[str] = word_tokenize(
-                    content,
-                    language=settings.NLTK_LANGUAGE,
-                )
-
-                meaningful_words = []
-                for word in words:
-                    # Skip stop words
-                    # These are words like "a", "and", "the" which add little meaning
-                    if word in self._stop_words:
-                        continue
-                    # Stem the words
-                    # This reduces the words to their stems.
-                    # "amazement" returns "amaz"
-                    # "amaze" returns "amaz
-                    # "amazed" returns "amaz"
-                    meaningful_words.append(self._stemmer.stem(word))
-
-                return " ".join(meaningful_words)
-
+                self._stemmer = SnowballStemmer(settings.NLTK_LANGUAGE)
+                self._stop_words = frozenset(stopwords.words(settings.NLTK_LANGUAGE))
             except AttributeError:
+                logger.debug("Could not initialize NLTK for advanced text processing.")
+                return False
+        return True
+
+    def stem_and_skip_stop_words(self, words: list[str], *, shared_cache=True):
+        """
+        Reduce a list of words to their stem. Stop words are converted to empty strings.
+        :param words: the list of words to stem
+        """
+
+        def _stem_and_skip_stop_word(word: str):
+            """
+            Reduce a given word to its stem. If it's a stop word, return an empty string.
+            E.g. "amazement", "amaze" and "amazed" all return "amaz".
+            """
+            cached = self._stem_cache.get(word)
+            if cached is not None:
+                return cached
+            elif word in self._stop_words:
+                return ""
+            # Assumption: words that contain numbers are never stemmed
+            elif RE_DIGIT.search(word):
+                return word
+            else:
+                result = self._stemmer.stem(word)
+                self._stem_cache.set(word, result)
+                return result
+
+        if shared_cache:
+            self._stem_cache.load()
+
+        # Stem the words and skip stop words
+        result = " ".join(
+            filter(None, (_stem_and_skip_stop_word(w) for w in words)),
+        )
+        if shared_cache:
+            self._stem_cache.save()
+        return result
+
+    def preprocess_content(
+        self,
+        content: str,
+        *,
+        shared_cache=True,
+    ) -> str:
+        """
+        Process the contents of a document, distilling it down into
+        words which are meaningful to the content.
+
+        A stemmer cache is shared across workers with the parameter "shared_cache".
+        This is unnecessary when training the classifier.
+        """
+
+        # Lower case the document, reduce space,
+        # and keep only letters and digits.
+        content = " ".join(match.group().lower() for match in RE_WORD.finditer(content))
+
+        if ADVANCED_TEXT_PROCESSING_ENABLED:
+            from nltk.tokenize import word_tokenize
+
+            if not self._init_advanced_text_processing():
                 return content
+            # Tokenize
+            # This splits the content into tokens, roughly words
+            words = word_tokenize(content, language=settings.NLTK_LANGUAGE)
+            # Stem the words and skip stop words
+            content = self.stem_and_skip_stop_words(words, shared_cache=shared_cache)
 
         return content
 
+    def _get_vectorizer_cache_key(self, content: str):
+        hash = sha256(content.encode())
+        hash.update(
+            f"|{self.FORMAT_VERSION}|{settings.NLTK_LANGUAGE}|{settings.NLTK_ENABLED}|{self.data_vectorizer_hash}".encode(),
+        )
+        return f"vectorized_content_{hash.hexdigest()}"
+
+    def _vectorize(self, content: str):
+        key = self._get_vectorizer_cache_key(content)
+        serialized_result = read_cache.get(key)
+        if serialized_result is None:
+            result = self.data_vectorizer.transform([self.preprocess_content(content)])
+            read_cache.set(key, pickle.dumps(result), CACHE_5_MINUTES)
+        else:
+            read_cache.touch(key, CACHE_5_MINUTES)
+            result = pickle.loads(serialized_result)
+        return result
+
     def predict_correspondent(self, content: str) -> int | None:
         if self.correspondent_classifier:
-            X = self.data_vectorizer.transform([self.preprocess_content(content)])
+            X = self._vectorize(content)
             correspondent_id = self.correspondent_classifier.predict(X)
             if correspondent_id != -1:
                 return correspondent_id
@@ -432,7 +504,7 @@ class DocumentClassifier:
 
     def predict_document_type(self, content: str) -> int | None:
         if self.document_type_classifier:
-            X = self.data_vectorizer.transform([self.preprocess_content(content)])
+            X = self._vectorize(content)
             document_type_id = self.document_type_classifier.predict(X)
             if document_type_id != -1:
                 return document_type_id
@@ -445,7 +517,7 @@ class DocumentClassifier:
         from sklearn.utils.multiclass import type_of_target
 
         if self.tags_classifier:
-            X = self.data_vectorizer.transform([self.preprocess_content(content)])
+            X = self._vectorize(content)
             y = self.tags_classifier.predict(X)
             tags_ids = self.tags_binarizer.inverse_transform(y)[0]
             if type_of_target(y).startswith("multilabel"):
@@ -464,7 +536,7 @@ class DocumentClassifier:
 
     def predict_storage_path(self, content: str) -> int | None:
         if self.storage_path_classifier:
-            X = self.data_vectorizer.transform([self.preprocess_content(content)])
+            X = self._vectorize(content)
             storage_path_id = self.storage_path_classifier.predict(X)
             if storage_path_id != -1:
                 return storage_path_id
diff --git a/src/documents/file_handling.py b/src/documents/file_handling.py
index 3d1a643df..3a0ffd9fb 100644
--- a/src/documents/file_handling.py
+++ b/src/documents/file_handling.py
@@ -1,4 +1,5 @@
 import os
+from pathlib import Path
 
 from django.conf import settings
 
@@ -7,19 +8,15 @@ from documents.templating.filepath import validate_filepath_template_and_render
 from documents.templating.utils import convert_format_str_to_template_format
 
 
-def create_source_path_directory(source_path):
-    os.makedirs(os.path.dirname(source_path), exist_ok=True)
+def create_source_path_directory(source_path: Path) -> None:
+    source_path.parent.mkdir(parents=True, exist_ok=True)
 
 
-def delete_empty_directories(directory, root):
-    if not os.path.isdir(directory):
+def delete_empty_directories(directory: Path, root: Path) -> None:
+    if not directory.is_dir():
         return
 
-    # Go up in the directory hierarchy and try to delete all directories
-    directory = os.path.normpath(directory)
-    root = os.path.normpath(root)
-
-    if not directory.startswith(root + os.path.sep):
+    if not directory.is_relative_to(root):
         # don't do anything outside our originals folder.
 
         # append os.path.set so that we avoid these cases:
@@ -27,11 +24,12 @@ def delete_empty_directories(directory, root):
         #   root = /home/originals ("/" gets appended and startswith fails)
         return
 
+    # Go up in the directory hierarchy and try to delete all directories
     while directory != root:
-        if not os.listdir(directory):
+        if not list(directory.iterdir()):
             # it's empty
             try:
-                os.rmdir(directory)
+                directory.rmdir()
             except OSError:
                 # whatever. empty directories aren't that bad anyway.
                 return
@@ -40,10 +38,10 @@ def delete_empty_directories(directory, root):
             return
 
         # go one level up
-        directory = os.path.normpath(os.path.dirname(directory))
+        directory = directory.parent
 
 
-def generate_unique_filename(doc, *, archive_filename=False):
+def generate_unique_filename(doc, *, archive_filename=False) -> Path:
     """
     Generates a unique filename for doc in settings.ORIGINALS_DIR.
 
@@ -56,21 +54,32 @@ def generate_unique_filename(doc, *, archive_filename=False):
 
     """
     if archive_filename:
-        old_filename = doc.archive_filename
+        old_filename: Path | None = (
+            Path(doc.archive_filename) if doc.archive_filename else None
+        )
         root = settings.ARCHIVE_DIR
     else:
-        old_filename = doc.filename
+        old_filename = Path(doc.filename) if doc.filename else None
         root = settings.ORIGINALS_DIR
 
     # If generating archive filenames, try to make a name that is similar to
     # the original filename first.
 
     if archive_filename and doc.filename:
-        new_filename = os.path.splitext(doc.filename)[0] + ".pdf"
-        if new_filename == old_filename or not os.path.exists(
-            os.path.join(root, new_filename),
-        ):
-            return new_filename
+        # Generate the full path using the same logic as generate_filename
+        base_generated = generate_filename(doc, archive_filename=archive_filename)
+
+        # Try to create a simple PDF version based on the original filename
+        # but preserve any directory structure from the template
+        if str(base_generated.parent) != ".":
+            # Has directory structure, preserve it
+            simple_pdf_name = base_generated.parent / (Path(doc.filename).stem + ".pdf")
+        else:
+            # No directory structure
+            simple_pdf_name = Path(Path(doc.filename).stem + ".pdf")
+
+        if simple_pdf_name == old_filename or not (root / simple_pdf_name).exists():
+            return simple_pdf_name
 
     counter = 0
 
@@ -84,7 +93,7 @@ def generate_unique_filename(doc, *, archive_filename=False):
             # still the same as before.
             return new_filename
 
-        if os.path.exists(os.path.join(root, new_filename)):
+        if (root / new_filename).exists():
             counter += 1
         else:
             return new_filename
@@ -96,8 +105,8 @@ def generate_filename(
     counter=0,
     append_gpg=True,
     archive_filename=False,
-):
-    path = ""
+) -> Path:
+    base_path: Path | None = None
 
     def format_filename(document: Document, template_str: str) -> str | None:
         rendered_filename = validate_filepath_template_and_render(
@@ -134,17 +143,34 @@ def generate_filename(
 
     # If we have one, render it
     if filename_format is not None:
-        path = format_filename(doc, filename_format)
+        rendered_path: str | None = format_filename(doc, filename_format)
+        if rendered_path:
+            base_path = Path(rendered_path)
 
     counter_str = f"_{counter:02}" if counter else ""
     filetype_str = ".pdf" if archive_filename else doc.file_type
 
-    if path:
-        filename = f"{path}{counter_str}{filetype_str}"
+    if base_path:
+        # Split the path into directory and filename parts
+        directory = base_path.parent
+        # Use the full name (not just stem) as the base filename
+        base_filename = base_path.name
+
+        # Build the final filename with counter and filetype
+        final_filename = f"{base_filename}{counter_str}{filetype_str}"
+
+        # If we have a directory component, include it
+        if str(directory) != ".":
+            full_path = directory / final_filename
+        else:
+            full_path = Path(final_filename)
     else:
-        filename = f"{doc.pk:07}{counter_str}{filetype_str}"
+        # No template, use document ID
+        final_filename = f"{doc.pk:07}{counter_str}{filetype_str}"
+        full_path = Path(final_filename)
 
+    # Add GPG extension if needed
     if append_gpg and doc.storage_type == doc.STORAGE_TYPE_GPG:
-        filename += ".gpg"
+        full_path = full_path.with_suffix(full_path.suffix + ".gpg")
 
-    return filename
+    return full_path
diff --git a/src/documents/management/commands/document_exporter.py b/src/documents/management/commands/document_exporter.py
index 6dc89479e..88daeddf5 100644
--- a/src/documents/management/commands/document_exporter.py
+++ b/src/documents/management/commands/document_exporter.py
@@ -236,10 +236,7 @@ class Command(CryptMixin, BaseCommand):
                 # now make an archive in the original target, with all files stored
                 if self.zip_export and temp_dir is not None:
                     shutil.make_archive(
-                        os.path.join(
-                            self.original_target,
-                            options["zip_name"],
-                        ),
+                        self.original_target / options["zip_name"],
                         format="zip",
                         root_dir=temp_dir.name,
                     )
@@ -342,7 +339,7 @@ class Command(CryptMixin, BaseCommand):
                 )
 
             if self.split_manifest:
-                manifest_name = Path(base_name + "-manifest.json")
+                manifest_name = base_name.with_name(f"{base_name.stem}-manifest.json")
                 if self.use_folder_prefix:
                     manifest_name = Path("json") / manifest_name
                 manifest_name = (self.target / manifest_name).resolve()
@@ -416,7 +413,7 @@ class Command(CryptMixin, BaseCommand):
                     else:
                         item.unlink()
 
-    def generate_base_name(self, document: Document) -> str:
+    def generate_base_name(self, document: Document) -> Path:
         """
         Generates a unique name for the document, one which hasn't already been exported (or will be)
         """
@@ -436,12 +433,12 @@ class Command(CryptMixin, BaseCommand):
                 break
             else:
                 filename_counter += 1
-        return base_name
+        return Path(base_name)
 
     def generate_document_targets(
         self,
         document: Document,
-        base_name: str,
+        base_name: Path,
         document_dict: dict,
     ) -> tuple[Path, Path | None, Path | None]:
         """
@@ -449,25 +446,25 @@ class Command(CryptMixin, BaseCommand):
         """
         original_name = base_name
         if self.use_folder_prefix:
-            original_name = os.path.join("originals", original_name)
-        original_target = (self.target / Path(original_name)).resolve()
-        document_dict[EXPORTER_FILE_NAME] = original_name
+            original_name = Path("originals") / original_name
+        original_target = (self.target / original_name).resolve()
+        document_dict[EXPORTER_FILE_NAME] = str(original_name)
 
         if not self.no_thumbnail:
-            thumbnail_name = base_name + "-thumbnail.webp"
+            thumbnail_name = base_name.parent / (base_name.stem + "-thumbnail.webp")
             if self.use_folder_prefix:
-                thumbnail_name = os.path.join("thumbnails", thumbnail_name)
-            thumbnail_target = (self.target / Path(thumbnail_name)).resolve()
-            document_dict[EXPORTER_THUMBNAIL_NAME] = thumbnail_name
+                thumbnail_name = Path("thumbnails") / thumbnail_name
+            thumbnail_target = (self.target / thumbnail_name).resolve()
+            document_dict[EXPORTER_THUMBNAIL_NAME] = str(thumbnail_name)
         else:
             thumbnail_target = None
 
         if not self.no_archive and document.has_archive_version:
-            archive_name = base_name + "-archive.pdf"
+            archive_name = base_name.parent / (base_name.stem + "-archive.pdf")
             if self.use_folder_prefix:
-                archive_name = os.path.join("archive", archive_name)
-            archive_target = (self.target / Path(archive_name)).resolve()
-            document_dict[EXPORTER_ARCHIVE_NAME] = archive_name
+                archive_name = Path("archive") / archive_name
+            archive_target = (self.target / archive_name).resolve()
+            document_dict[EXPORTER_ARCHIVE_NAME] = str(archive_name)
         else:
             archive_target = None
 
@@ -572,7 +569,7 @@ class Command(CryptMixin, BaseCommand):
         perform_copy = False
 
         if target.exists():
-            source_stat = os.stat(source)
+            source_stat = source.stat()
             target_stat = target.stat()
             if self.compare_checksums and source_checksum:
                 target_checksum = hashlib.md5(target.read_bytes()).hexdigest()
diff --git a/src/documents/migrations/0014_document_checksum.py b/src/documents/migrations/0014_document_checksum.py
index dac945e32..de256ced7 100644
--- a/src/documents/migrations/0014_document_checksum.py
+++ b/src/documents/migrations/0014_document_checksum.py
@@ -63,11 +63,11 @@ class Document:
             / "documents"
             / "originals"
             / f"{self.pk:07}.{self.file_type}.gpg"
-        ).as_posix()
+        )
 
     @property
     def source_file(self):
-        return Path(self.source_path).open("rb")
+        return self.source_path.open("rb")
 
     @property
     def file_name(self):
diff --git a/src/documents/signals/handlers.py b/src/documents/signals/handlers.py
index 0c20de95a..76df8d4e4 100644
--- a/src/documents/signals/handlers.py
+++ b/src/documents/signals/handlers.py
@@ -1,8 +1,8 @@
 from __future__ import annotations
 
 import logging
-import os
 import shutil
+from pathlib import Path
 from typing import TYPE_CHECKING
 
 import httpx
@@ -12,11 +12,13 @@ from celery.signals import before_task_publish
 from celery.signals import task_failure
 from celery.signals import task_postrun
 from celery.signals import task_prerun
+from celery.signals import worker_process_init
 from django.conf import settings
 from django.contrib.auth.models import Group
 from django.contrib.auth.models import User
 from django.db import DatabaseError
 from django.db import close_old_connections
+from django.db import connections
 from django.db import models
 from django.db.models import Q
 from django.dispatch import receiver
@@ -49,8 +51,6 @@ from documents.permissions import set_permissions_for_object
 from documents.templating.workflows import parse_w_workflow_placeholders
 
 if TYPE_CHECKING:
-    from pathlib import Path
-
     from documents.classifier import DocumentClassifier
     from documents.data_models import ConsumableDocument
     from documents.data_models import DocumentMetadataOverrides
@@ -327,15 +327,16 @@ def cleanup_document_deletion(sender, instance, **kwargs):
             # Find a non-conflicting filename in case a document with the same
             # name was moved to trash earlier
             counter = 0
-            old_filename = os.path.split(instance.source_path)[1]
-            (old_filebase, old_fileext) = os.path.splitext(old_filename)
+            old_filename = Path(instance.source_path).name
+            old_filebase = Path(old_filename).stem
+            old_fileext = Path(old_filename).suffix
 
             while True:
                 new_file_path = settings.EMPTY_TRASH_DIR / (
                     old_filebase + (f"_{counter:02}" if counter else "") + old_fileext
                 )
 
-                if os.path.exists(new_file_path):
+                if new_file_path.exists():
                     counter += 1
                 else:
                     break
@@ -359,26 +360,26 @@ def cleanup_document_deletion(sender, instance, **kwargs):
             files += (instance.source_path,)
 
         for filename in files:
-            if filename and os.path.isfile(filename):
+            if filename and filename.is_file():
                 try:
-                    os.unlink(filename)
+                    filename.unlink()
                     logger.debug(f"Deleted file {filename}.")
                 except OSError as e:
                     logger.warning(
                         f"While deleting document {instance!s}, the file "
                         f"{filename} could not be deleted: {e}",
                     )
-            elif filename and not os.path.isfile(filename):
+            elif filename and not filename.is_file():
                 logger.warning(f"Expected {filename} to exist, but it did not")
 
         delete_empty_directories(
-            os.path.dirname(instance.source_path),
+            Path(instance.source_path).parent,
             root=settings.ORIGINALS_DIR,
         )
 
         if instance.has_archive_version:
             delete_empty_directories(
-                os.path.dirname(instance.archive_path),
+                Path(instance.archive_path).parent,
                 root=settings.ARCHIVE_DIR,
             )
 
@@ -399,14 +400,14 @@ def update_filename_and_move_files(
     if isinstance(instance, CustomFieldInstance):
         instance = instance.document
 
-    def validate_move(instance, old_path, new_path):
-        if not os.path.isfile(old_path):
+    def validate_move(instance, old_path: Path, new_path: Path):
+        if not old_path.is_file():
             # Can't do anything if the old file does not exist anymore.
             msg = f"Document {instance!s}: File {old_path} doesn't exist."
             logger.fatal(msg)
             raise CannotMoveFilesException(msg)
 
-        if os.path.isfile(new_path):
+        if new_path.is_file():
             # Can't do anything if the new file already exists. Skip updating file.
             msg = f"Document {instance!s}: Cannot rename file since target path {new_path} already exists."
             logger.warning(msg)
@@ -434,16 +435,20 @@ def update_filename_and_move_files(
             old_filename = instance.filename
             old_source_path = instance.source_path
 
-            instance.filename = generate_unique_filename(instance)
+            # Need to convert to string to be able to save it to the db
+            instance.filename = str(generate_unique_filename(instance))
             move_original = old_filename != instance.filename
 
             old_archive_filename = instance.archive_filename
             old_archive_path = instance.archive_path
 
             if instance.has_archive_version:
-                instance.archive_filename = generate_unique_filename(
-                    instance,
-                    archive_filename=True,
+                # Need to convert to string to be able to save it to the db
+                instance.archive_filename = str(
+                    generate_unique_filename(
+                        instance,
+                        archive_filename=True,
+                    ),
                 )
 
                 move_archive = old_archive_filename != instance.archive_filename
@@ -485,11 +490,11 @@ def update_filename_and_move_files(
 
             # Try to move files to their original location.
             try:
-                if move_original and os.path.isfile(instance.source_path):
+                if move_original and instance.source_path.is_file():
                     logger.info("Restoring previous original path")
                     shutil.move(instance.source_path, old_source_path)
 
-                if move_archive and os.path.isfile(instance.archive_path):
+                if move_archive and instance.archive_path.is_file():
                     logger.info("Restoring previous archive path")
                     shutil.move(instance.archive_path, old_archive_path)
 
@@ -510,17 +515,15 @@ def update_filename_and_move_files(
 
         # finally, remove any empty sub folders. This will do nothing if
         # something has failed above.
-        if not os.path.isfile(old_source_path):
+        if not old_source_path.is_file():
             delete_empty_directories(
-                os.path.dirname(old_source_path),
+                Path(old_source_path).parent,
                 root=settings.ORIGINALS_DIR,
             )
 
-        if instance.has_archive_version and not os.path.isfile(
-            old_archive_path,
-        ):
+        if instance.has_archive_version and not old_archive_path.is_file():
             delete_empty_directories(
-                os.path.dirname(old_archive_path),
+                Path(old_archive_path).parent,
                 root=settings.ARCHIVE_DIR,
             )
 
@@ -1217,10 +1220,7 @@ def run_workflows(
                     )
             files = None
             if action.webhook.include_document:
-                with open(
-                    original_file,
-                    "rb",
-                ) as f:
+                with original_file.open("rb") as f:
                     files = {
                         "file": (
                             filename,
@@ -1439,3 +1439,18 @@ def task_failure_handler(
             task_instance.save()
     except Exception:  # pragma: no cover
         logger.exception("Updating PaperlessTask failed")
+
+
+@worker_process_init.connect
+def close_connection_pool_on_worker_init(**kwargs):
+    """
+    Close the DB connection pool for each Celery child process after it starts.
+
+    This is necessary because the parent process parse the Django configuration,
+    initializes connection pools then forks.
+
+    Closing these pools after forking ensures child processes have a valid connection.
+    """
+    for conn in connections.all(initialized_only=True):
+        if conn.alias == "default" and hasattr(conn, "pool") and conn.pool:
+            conn.close_pool()
diff --git a/src/documents/tests/samples/content.txt b/src/documents/tests/samples/content.txt
new file mode 100644
index 000000000..07666425c
--- /dev/null
+++ b/src/documents/tests/samples/content.txt
@@ -0,0 +1,34 @@
+Sample textual document content.
+Include as many characters as possible, to check the classifier's vectorization.
+
+Hey 00, this is "a" test0707 content.
+This is an example document — created on 2025-06-25.
+
+Digits: 0123456789
+Punctuation: . , ; : ! ? ' " ( ) [ ] { } — – …
+English text: The quick brown fox jumps over the lazy dog.
+English stop words: We’ve been doing it before.
+Accented Latin (diacritics): àâäæçéèêëîïôœùûüÿñ
+Arabic: لقد قام المترجم بعمل جيد
+Greek: Αλφα, Βήτα, Γάμμα, Δέλτα, Ωμέγα
+Cyrillic: Привет, как дела? Добро пожаловать!
+Chinese (Simplified): 你好，世界！今天的天气很好。
+Chinese (Traditional): 歡迎來到世界，今天天氣很好。
+Japanese (Kanji, Hiragana, Katakana): 東京へ行きます。カタカナ、ひらがな、漢字。
+Korean (Hangul): 안녕하세요. 오늘 날씨 어때요?
+Arabic: مرحبًا، كيف حالك؟
+Hebrew: שלום, מה שלומך?
+Emoji: 😀 🐍 📘 ✅ ©️ 🇺🇳
+Symbols: © ® ™ § ¶ † ‡ ∞ µ ∑ ∆ √
+Math: ∫₀^∞ x² dx = ∞, π ≈ 3.14159, ∇·E = ρ/ε₀
+Currency: 1$ € ¥ £ ₹
+Date formats: 25/06/2025, June 25, 2025, 2025年6月25日
+Quote in French: « Bonjour, ça va ? »
+Quote in German: „Guten Tag! Wie geht's?“
+Newline test:
+\r\n
+\r
+
+Tab\ttest\tspacing
+/ = +) ( []) ~ * #192 +33601010101 § ¤
+End of document.
diff --git a/src/documents/tests/samples/preprocessed_content.txt b/src/documents/tests/samples/preprocessed_content.txt
new file mode 100644
index 000000000..aac6beb7c
--- /dev/null
+++ b/src/documents/tests/samples/preprocessed_content.txt
@@ -0,0 +1 @@
+sample textual document content include as many characters as possible to check the classifier s vectorization hey 00 this is a test0707 content this is an example document created on 2025 06 25 digits 0123456789 punctuation english text the quick brown fox jumps over the lazy dog english stop words we ve been doing it before accented latin diacritics àâäæçéèêëîïôœùûüÿñ arabic لقد قام المترجم بعمل جيد greek αλφα βήτα γάμμα δέλτα ωμέγα cyrillic привет как дела добро пожаловать chinese simplified 你好 世界 今天的天气很好 chinese traditional 歡迎來到世界 今天天氣很好 japanese kanji hiragana katakana 東京へ行きます カタカナ ひらがな 漢字 korean hangul 안녕하세요 오늘 날씨 어때요 arabic مرحب ا كيف حالك hebrew שלום מה שלומך emoji symbols µ math ₀ x² dx π 3 14159 e ρ ε₀ currency 1 date formats 25 06 2025 june 25 2025 2025年6月25日 quote in french bonjour ça va quote in german guten tag wie geht s newline test r n r tab ttest tspacing 192 33601010101 end of document
diff --git a/src/documents/tests/samples/preprocessed_content_advanced.txt b/src/documents/tests/samples/preprocessed_content_advanced.txt
new file mode 100644
index 000000000..c67627c2b
--- /dev/null
+++ b/src/documents/tests/samples/preprocessed_content_advanced.txt
@@ -0,0 +1 @@
+sampl textual document content includ mani charact possibl check classifi vector hey 00 test0707 content exampl document creat 2025 06 25 digit 0123456789 punctuat english text quick brown fox jump lazi dog english stop word accent latin diacrit àâäæçéèêëîïôœùûüÿñ arab لقد قام المترجم بعمل جيد greek αλφα βήτα γάμμα δέλτα ωμέγα cyril привет как дела добро пожаловать chines simplifi 你好 世界 今天的天气很好 chines tradit 歡迎來到世界 今天天氣很好 japanes kanji hiragana katakana 東京へ行きます カタカナ ひらがな 漢字 korean hangul 안녕하세요 오늘 날씨 어때요 arab مرحب ا كيف حالك hebrew שלום מה שלומך emoji symbol µ math ₀ x² dx π 3 14159 e ρ ε₀ currenc 1 date format 25 06 2025 june 25 2025 2025年6月25日 quot french bonjour ça va quot german guten tag wie geht newlin test r n r tab ttest tspace 192 33601010101 end document
diff --git a/src/documents/tests/test_caching.py b/src/documents/tests/test_caching.py
new file mode 100644
index 000000000..4c8d2ead0
--- /dev/null
+++ b/src/documents/tests/test_caching.py
@@ -0,0 +1,45 @@
+import pickle
+
+from documents.caching import StoredLRUCache
+
+
+def test_lru_cache_entries():
+    CACHE_TTL = 1
+    # LRU cache with a capacity of 2 elements
+    cache = StoredLRUCache("test_lru_cache_key", 2, backend_ttl=CACHE_TTL)
+    cache.set(1, 1)
+    cache.set(2, 2)
+    assert cache.get(2) == 2
+    assert cache.get(1) == 1
+
+    # The oldest entry (2) should be removed
+    cache.set(3, 3)
+    assert cache.get(3) == 3
+    assert not cache.get(2)
+    assert cache.get(1) == 1
+
+    # Save the cache, restore it and check it overwrites the current cache in memory
+    cache.save()
+    cache.set(4, 4)
+    assert not cache.get(3)
+    cache.load()
+    assert not cache.get(4)
+    assert cache.get(3) == 3
+    assert cache.get(1) == 1
+
+
+def test_stored_lru_cache_key_ttl(mocker):
+    mock_backend = mocker.Mock()
+    cache = StoredLRUCache("test_key", backend=mock_backend, backend_ttl=321)
+
+    # Simulate storing values
+    cache.set("x", "X")
+    cache.set("y", "Y")
+    cache.save()
+
+    # Assert backend.set was called with pickled data, key and TTL
+    mock_backend.set.assert_called_once()
+    key, data, timeout = mock_backend.set.call_args[0]
+    assert key == "test_key"
+    assert timeout == 321
+    assert pickle.loads(data) == {"x": "X", "y": "Y"}
diff --git a/src/documents/tests/test_classifier.py b/src/documents/tests/test_classifier.py
index d1bc8e04f..b1317f70b 100644
--- a/src/documents/tests/test_classifier.py
+++ b/src/documents/tests/test_classifier.py
@@ -21,7 +21,7 @@ from documents.models import Tag
 from documents.tests.utils import DirectoriesMixin
 
 
-def dummy_preprocess(content: str):
+def dummy_preprocess(content: str, **kwargs):
     """
     Simpler, faster pre-processing for testing purposes
     """
@@ -223,24 +223,47 @@ class TestClassifier(DirectoriesMixin, TestCase):
         self.generate_test_data()
         self.classifier.train()
 
-        self.assertEqual(
-            self.classifier.predict_correspondent(self.doc1.content),
-            self.c1.pk,
-        )
-        self.assertEqual(self.classifier.predict_correspondent(self.doc2.content), None)
-        self.assertListEqual(
-            self.classifier.predict_tags(self.doc1.content),
-            [self.t1.pk],
-        )
-        self.assertListEqual(
-            self.classifier.predict_tags(self.doc2.content),
-            [self.t1.pk, self.t3.pk],
-        )
-        self.assertEqual(
-            self.classifier.predict_document_type(self.doc1.content),
-            self.dt.pk,
-        )
-        self.assertEqual(self.classifier.predict_document_type(self.doc2.content), None)
+        with (
+            mock.patch.object(
+                self.classifier.data_vectorizer,
+                "transform",
+                wraps=self.classifier.data_vectorizer.transform,
+            ) as mock_transform,
+            mock.patch.object(
+                self.classifier,
+                "preprocess_content",
+                wraps=self.classifier.preprocess_content,
+            ) as mock_preprocess_content,
+        ):
+            self.assertEqual(
+                self.classifier.predict_correspondent(self.doc1.content),
+                self.c1.pk,
+            )
+            self.assertEqual(
+                self.classifier.predict_correspondent(self.doc2.content),
+                None,
+            )
+            self.assertListEqual(
+                self.classifier.predict_tags(self.doc1.content),
+                [self.t1.pk],
+            )
+            self.assertListEqual(
+                self.classifier.predict_tags(self.doc2.content),
+                [self.t1.pk, self.t3.pk],
+            )
+            self.assertEqual(
+                self.classifier.predict_document_type(self.doc1.content),
+                self.dt.pk,
+            )
+            self.assertEqual(
+                self.classifier.predict_document_type(self.doc2.content),
+                None,
+            )
+
+            # Check that the classifier vectorized content and text preprocessing has been cached
+            # It should be called once per document (doc1 and doc2)
+            self.assertEqual(mock_preprocess_content.call_count, 2)
+            self.assertEqual(mock_transform.call_count, 2)
 
     def test_no_retrain_if_no_change(self):
         """
@@ -694,3 +717,67 @@ class TestClassifier(DirectoriesMixin, TestCase):
         mock_load.side_effect = Exception()
         with self.assertRaises(Exception):
             load_classifier(raise_exception=True)
+
+
+def test_preprocess_content():
+    """
+    GIVEN:
+        - Advanced text processing is enabled (default)
+    WHEN:
+        - Classifier preprocesses a document's content
+    THEN:
+        - Processed content matches the expected output (stemmed words)
+    """
+    with (Path(__file__).parent / "samples" / "content.txt").open("r") as f:
+        content = f.read()
+    with (Path(__file__).parent / "samples" / "preprocessed_content_advanced.txt").open(
+        "r",
+    ) as f:
+        expected_preprocess_content = f.read().rstrip()
+    classifier = DocumentClassifier()
+    result = classifier.preprocess_content(content)
+    assert result == expected_preprocess_content
+
+
+def test_preprocess_content_nltk_disabled():
+    """
+    GIVEN:
+        - Advanced text processing is disabled
+    WHEN:
+        - Classifier preprocesses a document's content
+    THEN:
+        - Processed content matches the expected output (unstemmed words)
+    """
+    with (Path(__file__).parent / "samples" / "content.txt").open("r") as f:
+        content = f.read()
+    with (Path(__file__).parent / "samples" / "preprocessed_content.txt").open(
+        "r",
+    ) as f:
+        expected_preprocess_content = f.read().rstrip()
+    classifier = DocumentClassifier()
+    with mock.patch("documents.classifier.ADVANCED_TEXT_PROCESSING_ENABLED", new=False):
+        result = classifier.preprocess_content(content)
+    assert result == expected_preprocess_content
+
+
+def test_preprocess_content_nltk_load_fail(mocker):
+    """
+    GIVEN:
+        - NLTK stop words fail to load
+    WHEN:
+        - Classifier preprocesses a document's content
+    THEN:
+        - Processed content matches the expected output (unstemmed words)
+    """
+    _module = mocker.MagicMock(name="nltk_corpus_mock")
+    _module.stopwords.words.side_effect = AttributeError()
+    mocker.patch.dict("sys.modules", {"nltk.corpus": _module})
+    classifier = DocumentClassifier()
+    with (Path(__file__).parent / "samples" / "content.txt").open("r") as f:
+        content = f.read()
+    with (Path(__file__).parent / "samples" / "preprocessed_content.txt").open(
+        "r",
+    ) as f:
+        expected_preprocess_content = f.read().rstrip()
+    result = classifier.preprocess_content(content)
+    assert result == expected_preprocess_content
diff --git a/src/documents/tests/test_document_model.py b/src/documents/tests/test_document_model.py
index 45c8441f6..87ebdb561 100644
--- a/src/documents/tests/test_document_model.py
+++ b/src/documents/tests/test_document_model.py
@@ -41,11 +41,9 @@ class TestDocument(TestCase):
         Path(file_path).touch()
         Path(thumb_path).touch()
 
-        with mock.patch("documents.signals.handlers.os.unlink") as mock_unlink:
+        with mock.patch("documents.signals.handlers.Path.unlink") as mock_unlink:
             document.delete()
             empty_trash([document.pk])
-            mock_unlink.assert_any_call(file_path)
-            mock_unlink.assert_any_call(thumb_path)
             self.assertEqual(mock_unlink.call_count, 2)
 
     def test_document_soft_delete(self):
@@ -63,7 +61,7 @@ class TestDocument(TestCase):
         Path(file_path).touch()
         Path(thumb_path).touch()
 
-        with mock.patch("documents.signals.handlers.os.unlink") as mock_unlink:
+        with mock.patch("documents.signals.handlers.Path.unlink") as mock_unlink:
             document.delete()
             self.assertEqual(mock_unlink.call_count, 0)
 
diff --git a/src/documents/tests/test_file_handling.py b/src/documents/tests/test_file_handling.py
index d5bfa7184..d879137b9 100644
--- a/src/documents/tests/test_file_handling.py
+++ b/src/documents/tests/test_file_handling.py
@@ -34,12 +34,12 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
         document.storage_type = Document.STORAGE_TYPE_UNENCRYPTED
         document.save()
 
-        self.assertEqual(generate_filename(document), f"{document.pk:07d}.pdf")
+        self.assertEqual(generate_filename(document), Path(f"{document.pk:07d}.pdf"))
 
         document.storage_type = Document.STORAGE_TYPE_GPG
         self.assertEqual(
             generate_filename(document),
-            f"{document.pk:07d}.pdf.gpg",
+            Path(f"{document.pk:07d}.pdf.gpg"),
         )
 
     @override_settings(FILENAME_FORMAT="{correspondent}/{correspondent}")
@@ -58,12 +58,12 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
         document.filename = generate_filename(document)
 
         # Ensure that filename is properly generated
-        self.assertEqual(document.filename, "none/none.pdf")
+        self.assertEqual(document.filename, Path("none/none.pdf"))
 
         # Enable encryption and check again
         document.storage_type = Document.STORAGE_TYPE_GPG
         document.filename = generate_filename(document)
-        self.assertEqual(document.filename, "none/none.pdf.gpg")
+        self.assertEqual(document.filename, Path("none/none.pdf.gpg"))
 
         document.save()
 
@@ -96,7 +96,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
 
         # Ensure that filename is properly generated
         document.filename = generate_filename(document)
-        self.assertEqual(document.filename, "none/none.pdf")
+        self.assertEqual(document.filename, Path("none/none.pdf"))
         create_source_path_directory(document.source_path)
         document.source_path.touch()
 
@@ -137,7 +137,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
 
         # Ensure that filename is properly generated
         document.filename = generate_filename(document)
-        self.assertEqual(document.filename, "none/none.pdf")
+        self.assertEqual(document.filename, Path("none/none.pdf"))
         create_source_path_directory(document.source_path)
         Path(document.source_path).touch()
 
@@ -247,7 +247,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
 
         # Ensure that filename is properly generated
         document.filename = generate_filename(document)
-        self.assertEqual(document.filename, "none/none.pdf")
+        self.assertEqual(document.filename, Path("none/none.pdf"))
 
         create_source_path_directory(document.source_path)
 
@@ -269,11 +269,11 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
         dt = DocumentType.objects.create(name="my_doc_type")
         d = Document.objects.create(title="the_doc", mime_type="application/pdf")
 
-        self.assertEqual(generate_filename(d), "none - the_doc.pdf")
+        self.assertEqual(generate_filename(d), Path("none - the_doc.pdf"))
 
         d.document_type = dt
 
-        self.assertEqual(generate_filename(d), "my_doc_type - the_doc.pdf")
+        self.assertEqual(generate_filename(d), Path("my_doc_type - the_doc.pdf"))
 
     @override_settings(FILENAME_FORMAT="{asn} - {title}")
     def test_asn(self):
@@ -289,8 +289,8 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
             archive_serial_number=None,
             checksum="B",
         )
-        self.assertEqual(generate_filename(d1), "652 - the_doc.pdf")
-        self.assertEqual(generate_filename(d2), "none - the_doc.pdf")
+        self.assertEqual(generate_filename(d1), Path("652 - the_doc.pdf"))
+        self.assertEqual(generate_filename(d2), Path("none - the_doc.pdf"))
 
     @override_settings(FILENAME_FORMAT="{title} {tag_list}")
     def test_tag_list(self):
@@ -298,7 +298,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
         doc.tags.create(name="tag2")
         doc.tags.create(name="tag1")
 
-        self.assertEqual(generate_filename(doc), "doc1 tag1,tag2.pdf")
+        self.assertEqual(generate_filename(doc), Path("doc1 tag1,tag2.pdf"))
 
         doc = Document.objects.create(
             title="doc2",
@@ -306,7 +306,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
             mime_type="application/pdf",
         )
 
-        self.assertEqual(generate_filename(doc), "doc2.pdf")
+        self.assertEqual(generate_filename(doc), Path("doc2.pdf"))
 
     @override_settings(FILENAME_FORMAT="//etc/something/{title}")
     def test_filename_relative(self):
@@ -330,11 +330,11 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
             created=d1,
         )
 
-        self.assertEqual(generate_filename(doc1), "2020-03-06.pdf")
+        self.assertEqual(generate_filename(doc1), Path("2020-03-06.pdf"))
 
         doc1.created = datetime.date(2020, 11, 16)
 
-        self.assertEqual(generate_filename(doc1), "2020-11-16.pdf")
+        self.assertEqual(generate_filename(doc1), Path("2020-11-16.pdf"))
 
     @override_settings(
         FILENAME_FORMAT="{added_year}-{added_month}-{added_day}",
@@ -347,11 +347,11 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
             added=d1,
         )
 
-        self.assertEqual(generate_filename(doc1), "232-01-09.pdf")
+        self.assertEqual(generate_filename(doc1), Path("232-01-09.pdf"))
 
         doc1.added = timezone.make_aware(datetime.datetime(2020, 11, 16, 1, 1, 1))
 
-        self.assertEqual(generate_filename(doc1), "2020-11-16.pdf")
+        self.assertEqual(generate_filename(doc1), Path("2020-11-16.pdf"))
 
     @override_settings(
         FILENAME_FORMAT="{correspondent}/{correspondent}/{correspondent}",
@@ -389,11 +389,11 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
         document.mime_type = "application/pdf"
         document.storage_type = Document.STORAGE_TYPE_UNENCRYPTED
 
-        self.assertEqual(generate_filename(document), "0000001.pdf")
+        self.assertEqual(generate_filename(document), Path("0000001.pdf"))
 
         document.pk = 13579
 
-        self.assertEqual(generate_filename(document), "0013579.pdf")
+        self.assertEqual(generate_filename(document), Path("0013579.pdf"))
 
     @override_settings(FILENAME_FORMAT=None)
     def test_format_none(self):
@@ -402,7 +402,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
         document.mime_type = "application/pdf"
         document.storage_type = Document.STORAGE_TYPE_UNENCRYPTED
 
-        self.assertEqual(generate_filename(document), "0000001.pdf")
+        self.assertEqual(generate_filename(document), Path("0000001.pdf"))
 
     def test_try_delete_empty_directories(self):
         # Create our working directory
@@ -428,7 +428,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
         document.mime_type = "application/pdf"
         document.storage_type = Document.STORAGE_TYPE_UNENCRYPTED
 
-        self.assertEqual(generate_filename(document), "0000001.pdf")
+        self.assertEqual(generate_filename(document), Path("0000001.pdf"))
 
     @override_settings(FILENAME_FORMAT="{created__year}")
     def test_invalid_format_key(self):
@@ -437,7 +437,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
         document.mime_type = "application/pdf"
         document.storage_type = Document.STORAGE_TYPE_UNENCRYPTED
 
-        self.assertEqual(generate_filename(document), "0000001.pdf")
+        self.assertEqual(generate_filename(document), Path("0000001.pdf"))
 
     @override_settings(FILENAME_FORMAT="{title}")
     def test_duplicates(self):
@@ -564,7 +564,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
             value_select="abc123",
         )
 
-        self.assertEqual(generate_filename(doc), "document_apple.pdf")
+        self.assertEqual(generate_filename(doc), Path("document_apple.pdf"))
 
         # handler should not have been called
         self.assertEqual(m.call_count, 0)
@@ -576,7 +576,7 @@ class TestFileHandling(DirectoriesMixin, FileSystemAssertsMixin, TestCase):
             ],
         }
         cf.save()
-        self.assertEqual(generate_filename(doc), "document_aubergine.pdf")
+        self.assertEqual(generate_filename(doc), Path("document_aubergine.pdf"))
         # handler should have been called
         self.assertEqual(m.call_count, 1)
 
@@ -897,7 +897,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             pk=1,
             checksum="1",
         )
-        self.assertEqual(generate_filename(doc), "This. is the title.pdf")
+        self.assertEqual(generate_filename(doc), Path("This. is the title.pdf"))
 
         doc = Document.objects.create(
             title="my\\invalid/../title:yay",
@@ -905,7 +905,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             pk=2,
             checksum="2",
         )
-        self.assertEqual(generate_filename(doc), "my-invalid-..-title-yay.pdf")
+        self.assertEqual(generate_filename(doc), Path("my-invalid-..-title-yay.pdf"))
 
     @override_settings(FILENAME_FORMAT="{created}")
     def test_date(self):
@@ -916,7 +916,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             pk=2,
             checksum="2",
         )
-        self.assertEqual(generate_filename(doc), "2020-05-21.pdf")
+        self.assertEqual(generate_filename(doc), Path("2020-05-21.pdf"))
 
     def test_dynamic_path(self):
         """
@@ -935,7 +935,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             checksum="2",
             storage_path=StoragePath.objects.create(path="TestFolder/{{created}}"),
         )
-        self.assertEqual(generate_filename(doc), "TestFolder/2020-06-25.pdf")
+        self.assertEqual(generate_filename(doc), Path("TestFolder/2020-06-25.pdf"))
 
     def test_dynamic_path_with_none(self):
         """
@@ -956,7 +956,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             checksum="2",
             storage_path=StoragePath.objects.create(path="{{asn}} - {{created}}"),
         )
-        self.assertEqual(generate_filename(doc), "none - 2020-06-25.pdf")
+        self.assertEqual(generate_filename(doc), Path("none - 2020-06-25.pdf"))
 
     @override_settings(
         FILENAME_FORMAT_REMOVE_NONE=True,
@@ -984,7 +984,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             checksum="2",
             storage_path=sp,
         )
-        self.assertEqual(generate_filename(doc), "TestFolder/2020-06-25.pdf")
+        self.assertEqual(generate_filename(doc), Path("TestFolder/2020-06-25.pdf"))
 
         # Special case, undefined variable, then defined at the start of the template
         # This could lead to an absolute path after we remove the leading -none-, but leave the leading /
@@ -993,7 +993,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             "{{ owner_username }}/{{ created_year }}/{{ correspondent }}/{{ title }}"
         )
         sp.save()
-        self.assertEqual(generate_filename(doc), "2020/does not matter.pdf")
+        self.assertEqual(generate_filename(doc), Path("2020/does not matter.pdf"))
 
     def test_multiple_doc_paths(self):
         """
@@ -1028,8 +1028,14 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             ),
         )
 
-        self.assertEqual(generate_filename(doc_a), "ThisIsAFolder/4/2020-06-25.pdf")
-        self.assertEqual(generate_filename(doc_b), "SomeImportantNone/2020-07-25.pdf")
+        self.assertEqual(
+            generate_filename(doc_a),
+            Path("ThisIsAFolder/4/2020-06-25.pdf"),
+        )
+        self.assertEqual(
+            generate_filename(doc_b),
+            Path("SomeImportantNone/2020-07-25.pdf"),
+        )
 
     @override_settings(
         FILENAME_FORMAT=None,
@@ -1064,8 +1070,11 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             ),
         )
 
-        self.assertEqual(generate_filename(doc_a), "0000002.pdf")
-        self.assertEqual(generate_filename(doc_b), "SomeImportantNone/2020-07-25.pdf")
+        self.assertEqual(generate_filename(doc_a), Path("0000002.pdf"))
+        self.assertEqual(
+            generate_filename(doc_b),
+            Path("SomeImportantNone/2020-07-25.pdf"),
+        )
 
     @override_settings(
         FILENAME_FORMAT="{created_year_short}/{created_month_name_short}/{created_month_name}/{title}",
@@ -1078,7 +1087,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             pk=2,
             checksum="2",
         )
-        self.assertEqual(generate_filename(doc), "89/Dec/December/The Title.pdf")
+        self.assertEqual(generate_filename(doc), Path("89/Dec/December/The Title.pdf"))
 
     @override_settings(
         FILENAME_FORMAT="{added_year_short}/{added_month_name}/{added_month_name_short}/{title}",
@@ -1091,7 +1100,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             pk=2,
             checksum="2",
         )
-        self.assertEqual(generate_filename(doc), "84/August/Aug/The Title.pdf")
+        self.assertEqual(generate_filename(doc), Path("84/August/Aug/The Title.pdf"))
 
     @override_settings(
         FILENAME_FORMAT="{owner_username}/{title}",
@@ -1124,8 +1133,8 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             checksum="3",
         )
 
-        self.assertEqual(generate_filename(owned_doc), "user1/The Title.pdf")
-        self.assertEqual(generate_filename(no_owner_doc), "none/does matter.pdf")
+        self.assertEqual(generate_filename(owned_doc), Path("user1/The Title.pdf"))
+        self.assertEqual(generate_filename(no_owner_doc), Path("none/does matter.pdf"))
 
     @override_settings(
         FILENAME_FORMAT="{original_name}",
@@ -1171,17 +1180,20 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
             original_filename="logs.txt",
         )
 
-        self.assertEqual(generate_filename(doc_with_original), "someepdf.pdf")
+        self.assertEqual(generate_filename(doc_with_original), Path("someepdf.pdf"))
 
         self.assertEqual(
             generate_filename(tricky_with_original),
-            "some pdf with spaces and stuff.pdf",
+            Path("some pdf with spaces and stuff.pdf"),
         )
 
-        self.assertEqual(generate_filename(no_original), "none.pdf")
+        self.assertEqual(generate_filename(no_original), Path("none.pdf"))
 
-        self.assertEqual(generate_filename(text_doc), "logs.txt")
-        self.assertEqual(generate_filename(text_doc, archive_filename=True), "logs.pdf")
+        self.assertEqual(generate_filename(text_doc), Path("logs.txt"))
+        self.assertEqual(
+            generate_filename(text_doc, archive_filename=True),
+            Path("logs.pdf"),
+        )
 
     @override_settings(
         FILENAME_FORMAT="XX{correspondent}/{title}",
@@ -1206,7 +1218,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
 
         # Ensure that filename is properly generated
         document.filename = generate_filename(document)
-        self.assertEqual(document.filename, "XX/doc1.pdf")
+        self.assertEqual(document.filename, Path("XX/doc1.pdf"))
 
     def test_complex_template_strings(self):
         """
@@ -1244,19 +1256,19 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
 
         self.assertEqual(
             generate_filename(doc_a),
-            "somepath/some where/2020-06-25/Does Matter.pdf",
+            Path("somepath/some where/2020-06-25/Does Matter.pdf"),
         )
         doc_a.checksum = "5"
 
         self.assertEqual(
             generate_filename(doc_a),
-            "somepath/2024-10-01/Does Matter.pdf",
+            Path("somepath/2024-10-01/Does Matter.pdf"),
         )
 
         sp.path = "{{ document.title|lower }}{{ document.archive_serial_number - 2 }}"
         sp.save()
 
-        self.assertEqual(generate_filename(doc_a), "does matter23.pdf")
+        self.assertEqual(generate_filename(doc_a), Path("does matter23.pdf"))
 
         sp.path = """
                  somepath/
@@ -1275,13 +1287,13 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         sp.save()
         self.assertEqual(
             generate_filename(doc_a),
-            "somepath/asn-000-200/Does Matter/Does Matter.pdf",
+            Path("somepath/asn-000-200/Does Matter/Does Matter.pdf"),
         )
         doc_a.archive_serial_number = 301
         doc_a.save()
         self.assertEqual(
             generate_filename(doc_a),
-            "somepath/asn-201-400/asn-3xx/Does Matter.pdf",
+            Path("somepath/asn-201-400/asn-3xx/Does Matter.pdf"),
         )
 
     @override_settings(
@@ -1310,7 +1322,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         with self.assertLogs(level=logging.WARNING) as capture:
             self.assertEqual(
                 generate_filename(doc_a),
-                "0000002.pdf",
+                Path("0000002.pdf"),
             )
 
             self.assertEqual(len(capture.output), 1)
@@ -1345,7 +1357,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         with self.assertLogs(level=logging.WARNING) as capture:
             self.assertEqual(
                 generate_filename(doc_a),
-                "0000002.pdf",
+                Path("0000002.pdf"),
             )
 
             self.assertEqual(len(capture.output), 1)
@@ -1413,7 +1425,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc_a),
-                "invoices/1234.pdf",
+                Path("invoices/1234.pdf"),
             )
 
         with override_settings(
@@ -1427,7 +1439,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc_a),
-                "Some Title_ChoiceOne.pdf",
+                Path("Some Title_ChoiceOne.pdf"),
             )
 
             # Check for handling Nones well
@@ -1436,7 +1448,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
 
             self.assertEqual(
                 generate_filename(doc_a),
-                "Some Title_Default Value.pdf",
+                Path("Some Title_Default Value.pdf"),
             )
 
         cf.name = "Invoice Number"
@@ -1449,7 +1461,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc_a),
-                "invoices/4567.pdf",
+                Path("invoices/4567.pdf"),
             )
 
         with override_settings(
@@ -1457,7 +1469,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc_a),
-                "invoices/0.pdf",
+                Path("invoices/0.pdf"),
             )
 
     def test_datetime_filter(self):
@@ -1496,7 +1508,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc_a),
-                "2020/Some Title.pdf",
+                Path("2020/Some Title.pdf"),
             )
 
         with override_settings(
@@ -1504,7 +1516,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc_a),
-                "2020-06-25/Some Title.pdf",
+                Path("2020-06-25/Some Title.pdf"),
             )
 
         with override_settings(
@@ -1512,7 +1524,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc_a),
-                "2024-10-01/Some Title.pdf",
+                Path("2024-10-01/Some Title.pdf"),
             )
 
     def test_slugify_filter(self):
@@ -1539,7 +1551,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc),
-                "some-title-with-special-characters.pdf",
+                Path("some-title-with-special-characters.pdf"),
             )
 
         # Test with correspondent name containing spaces and special chars
@@ -1553,7 +1565,7 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc),
-                "johns-office-workplace/some-title-with-special-characters.pdf",
+                Path("johns-office-workplace/some-title-with-special-characters.pdf"),
             )
 
         # Test with custom fields
@@ -1572,5 +1584,5 @@ class TestFilenameGeneration(DirectoriesMixin, TestCase):
         ):
             self.assertEqual(
                 generate_filename(doc),
-                "brussels-belgium/some-title-with-special-characters.pdf",
+                Path("brussels-belgium/some-title-with-special-characters.pdf"),
             )
diff --git a/src/documents/tests/test_management_exporter.py b/src/documents/tests/test_management_exporter.py
index 292151f70..68d204765 100644
--- a/src/documents/tests/test_management_exporter.py
+++ b/src/documents/tests/test_management_exporter.py
@@ -209,7 +209,7 @@ class TestExportImport(
             4,
         )
 
-        self.assertIsFile((self.target / "manifest.json").as_posix())
+        self.assertIsFile(self.target / "manifest.json")
 
         self.assertEqual(
             self._get_document_from_manifest(manifest, self.d1.id)["fields"]["title"],
@@ -235,9 +235,7 @@ class TestExportImport(
                 ).as_posix()
                 self.assertIsFile(fname)
                 self.assertIsFile(
-                    (
-                        self.target / element[document_exporter.EXPORTER_THUMBNAIL_NAME]
-                    ).as_posix(),
+                    self.target / element[document_exporter.EXPORTER_THUMBNAIL_NAME],
                 )
 
                 with Path(fname).open("rb") as f:
@@ -252,7 +250,7 @@ class TestExportImport(
                 if document_exporter.EXPORTER_ARCHIVE_NAME in element:
                     fname = (
                         self.target / element[document_exporter.EXPORTER_ARCHIVE_NAME]
-                    ).as_posix()
+                    )
                     self.assertIsFile(fname)
 
                     with Path(fname).open("rb") as f:
@@ -312,7 +310,7 @@ class TestExportImport(
         )
 
         self._do_export()
-        self.assertIsFile((self.target / "manifest.json").as_posix())
+        self.assertIsFile(self.target / "manifest.json")
 
         st_mtime_1 = (self.target / "manifest.json").stat().st_mtime
 
@@ -322,7 +320,7 @@ class TestExportImport(
             self._do_export()
             m.assert_not_called()
 
-        self.assertIsFile((self.target / "manifest.json").as_posix())
+        self.assertIsFile(self.target / "manifest.json")
         st_mtime_2 = (self.target / "manifest.json").stat().st_mtime
 
         Path(self.d1.source_path).touch()
@@ -334,7 +332,7 @@ class TestExportImport(
             self.assertEqual(m.call_count, 1)
 
         st_mtime_3 = (self.target / "manifest.json").stat().st_mtime
-        self.assertIsFile((self.target / "manifest.json").as_posix())
+        self.assertIsFile(self.target / "manifest.json")
 
         self.assertNotEqual(st_mtime_1, st_mtime_2)
         self.assertNotEqual(st_mtime_2, st_mtime_3)
@@ -352,7 +350,7 @@ class TestExportImport(
 
         self._do_export()
 
-        self.assertIsFile((self.target / "manifest.json").as_posix())
+        self.assertIsFile(self.target / "manifest.json")
 
         with mock.patch(
             "documents.management.commands.document_exporter.copy_file_with_basic_stats",
@@ -360,7 +358,7 @@ class TestExportImport(
             self._do_export()
             m.assert_not_called()
 
-        self.assertIsFile((self.target / "manifest.json").as_posix())
+        self.assertIsFile(self.target / "manifest.json")
 
         self.d2.checksum = "asdfasdgf3"
         self.d2.save()
@@ -371,7 +369,7 @@ class TestExportImport(
             self._do_export(compare_checksums=True)
             self.assertEqual(m.call_count, 1)
 
-        self.assertIsFile((self.target / "manifest.json").as_posix())
+        self.assertIsFile(self.target / "manifest.json")
 
     def test_update_export_deleted_document(self):
         shutil.rmtree(Path(self.dirs.media_dir) / "documents")
@@ -385,7 +383,7 @@ class TestExportImport(
         self.assertTrue(len(manifest), 7)
         doc_from_manifest = self._get_document_from_manifest(manifest, self.d3.id)
         self.assertIsFile(
-            (self.target / doc_from_manifest[EXPORTER_FILE_NAME]).as_posix(),
+            str(self.target / doc_from_manifest[EXPORTER_FILE_NAME]),
         )
         self.d3.delete()
 
@@ -397,12 +395,12 @@ class TestExportImport(
             self.d3.id,
         )
         self.assertIsFile(
-            (self.target / doc_from_manifest[EXPORTER_FILE_NAME]).as_posix(),
+            self.target / doc_from_manifest[EXPORTER_FILE_NAME],
         )
 
         manifest = self._do_export(delete=True)
         self.assertIsNotFile(
-            (self.target / doc_from_manifest[EXPORTER_FILE_NAME]).as_posix(),
+            self.target / doc_from_manifest[EXPORTER_FILE_NAME],
         )
 
         self.assertTrue(len(manifest), 6)
@@ -416,20 +414,20 @@ class TestExportImport(
         )
 
         self._do_export(use_filename_format=True)
-        self.assertIsFile((self.target / "wow1" / "c.pdf").as_posix())
+        self.assertIsFile(self.target / "wow1" / "c.pdf")
 
-        self.assertIsFile((self.target / "manifest.json").as_posix())
+        self.assertIsFile(self.target / "manifest.json")
 
         self.d1.title = "new_title"
         self.d1.save()
         self._do_export(use_filename_format=True, delete=True)
-        self.assertIsNotFile((self.target / "wow1" / "c.pdf").as_posix())
-        self.assertIsNotDir((self.target / "wow1").as_posix())
-        self.assertIsFile((self.target / "new_title" / "c.pdf").as_posix())
-        self.assertIsFile((self.target / "manifest.json").as_posix())
-        self.assertIsFile((self.target / "wow2" / "none.pdf").as_posix())
+        self.assertIsNotFile(self.target / "wow1" / "c.pdf")
+        self.assertIsNotDir(self.target / "wow1")
+        self.assertIsFile(self.target / "new_title" / "c.pdf")
+        self.assertIsFile(self.target / "manifest.json")
+        self.assertIsFile(self.target / "wow2" / "none.pdf")
         self.assertIsFile(
-            (self.target / "wow2" / "none_01.pdf").as_posix(),
+            self.target / "wow2" / "none_01.pdf",
         )
 
     def test_export_missing_files(self):
diff --git a/src/documents/tests/test_migration_mime_type.py b/src/documents/tests/test_migration_mime_type.py
index f92381339..7805799fe 100644
--- a/src/documents/tests/test_migration_mime_type.py
+++ b/src/documents/tests/test_migration_mime_type.py
@@ -20,7 +20,7 @@ def source_path_before(self):
         if self.storage_type == STORAGE_TYPE_GPG:
             fname += ".gpg"
 
-    return (Path(settings.ORIGINALS_DIR) / fname).as_posix()
+    return Path(settings.ORIGINALS_DIR) / fname
 
 
 def file_type_after(self):
@@ -35,7 +35,7 @@ def source_path_after(doc):
         if doc.storage_type == STORAGE_TYPE_GPG:
             fname += ".gpg"  # pragma: no cover
 
-    return (Path(settings.ORIGINALS_DIR) / fname).as_posix()
+    return Path(settings.ORIGINALS_DIR) / fname
 
 
 @override_settings(PASSPHRASE="test")
diff --git a/src/locale/en_US/LC_MESSAGES/django.po b/src/locale/en_US/LC_MESSAGES/django.po
index e5ca72e2c..c3dc989b4 100644
--- a/src/locale/en_US/LC_MESSAGES/django.po
+++ b/src/locale/en_US/LC_MESSAGES/django.po
@@ -2,7 +2,7 @@ msgid ""
 msgstr ""
 "Project-Id-Version: paperless-ngx\n"
 "Report-Msgid-Bugs-To: \n"
-"POT-Creation-Date: 2025-07-08 21:14+0000\n"
+"POT-Creation-Date: 2025-08-02 12:55+0000\n"
 "PO-Revision-Date: 2022-02-17 04:17\n"
 "Last-Translator: \n"
 "Language-Team: English\n"
@@ -1645,147 +1645,147 @@ msgstr ""
 msgid "paperless application settings"
 msgstr ""
 
-#: paperless/settings.py:762
+#: paperless/settings.py:774
 msgid "English (US)"
 msgstr ""
 
-#: paperless/settings.py:763
+#: paperless/settings.py:775
 msgid "Arabic"
 msgstr ""
 
-#: paperless/settings.py:764
+#: paperless/settings.py:776
 msgid "Afrikaans"
 msgstr ""
 
-#: paperless/settings.py:765
+#: paperless/settings.py:777
 msgid "Belarusian"
 msgstr ""
 
-#: paperless/settings.py:766
+#: paperless/settings.py:778
 msgid "Bulgarian"
 msgstr ""
 
-#: paperless/settings.py:767
+#: paperless/settings.py:779
 msgid "Catalan"
 msgstr ""
 
-#: paperless/settings.py:768
+#: paperless/settings.py:780
 msgid "Czech"
 msgstr ""
 
-#: paperless/settings.py:769
+#: paperless/settings.py:781
 msgid "Danish"
 msgstr ""
 
-#: paperless/settings.py:770
+#: paperless/settings.py:782
 msgid "German"
 msgstr ""
 
-#: paperless/settings.py:771
+#: paperless/settings.py:783
 msgid "Greek"
 msgstr ""
 
-#: paperless/settings.py:772
+#: paperless/settings.py:784
 msgid "English (GB)"
 msgstr ""
 
-#: paperless/settings.py:773
+#: paperless/settings.py:785
 msgid "Spanish"
 msgstr ""
 
-#: paperless/settings.py:774
+#: paperless/settings.py:786
 msgid "Persian"
 msgstr ""
 
-#: paperless/settings.py:775
+#: paperless/settings.py:787
 msgid "Finnish"
 msgstr ""
 
-#: paperless/settings.py:776
+#: paperless/settings.py:788
 msgid "French"
 msgstr ""
 
-#: paperless/settings.py:777
+#: paperless/settings.py:789
 msgid "Hungarian"
 msgstr ""
 
-#: paperless/settings.py:778
+#: paperless/settings.py:790
 msgid "Italian"
 msgstr ""
 
-#: paperless/settings.py:779
+#: paperless/settings.py:791
 msgid "Japanese"
 msgstr ""
 
-#: paperless/settings.py:780
+#: paperless/settings.py:792
 msgid "Korean"
 msgstr ""
 
-#: paperless/settings.py:781
+#: paperless/settings.py:793
 msgid "Luxembourgish"
 msgstr ""
 
-#: paperless/settings.py:782
+#: paperless/settings.py:794
 msgid "Norwegian"
 msgstr ""
 
-#: paperless/settings.py:783
+#: paperless/settings.py:795
 msgid "Dutch"
 msgstr ""
 
-#: paperless/settings.py:784
+#: paperless/settings.py:796
 msgid "Polish"
 msgstr ""
 
-#: paperless/settings.py:785
+#: paperless/settings.py:797
 msgid "Portuguese (Brazil)"
 msgstr ""
 
-#: paperless/settings.py:786
+#: paperless/settings.py:798
 msgid "Portuguese"
 msgstr ""
 
-#: paperless/settings.py:787
+#: paperless/settings.py:799
 msgid "Romanian"
 msgstr ""
 
-#: paperless/settings.py:788
+#: paperless/settings.py:800
 msgid "Russian"
 msgstr ""
 
-#: paperless/settings.py:789
+#: paperless/settings.py:801
 msgid "Slovak"
 msgstr ""
 
-#: paperless/settings.py:790
+#: paperless/settings.py:802
 msgid "Slovenian"
 msgstr ""
 
-#: paperless/settings.py:791
+#: paperless/settings.py:803
 msgid "Serbian"
 msgstr ""
 
-#: paperless/settings.py:792
+#: paperless/settings.py:804
 msgid "Swedish"
 msgstr ""
 
-#: paperless/settings.py:793
+#: paperless/settings.py:805
 msgid "Turkish"
 msgstr ""
 
-#: paperless/settings.py:794
+#: paperless/settings.py:806
 msgid "Ukrainian"
 msgstr ""
 
-#: paperless/settings.py:795
+#: paperless/settings.py:807
 msgid "Vietnamese"
 msgstr ""
 
-#: paperless/settings.py:796
+#: paperless/settings.py:808
 msgid "Chinese Simplified"
 msgstr ""
 
-#: paperless/settings.py:797
+#: paperless/settings.py:809
 msgid "Chinese Traditional"
 msgstr ""
 
diff --git a/src/paperless/settings.py b/src/paperless/settings.py
index 08b781775..5d6d36422 100644
--- a/src/paperless/settings.py
+++ b/src/paperless/settings.py
@@ -704,6 +704,9 @@ def _parse_db_settings() -> dict:
         # Leave room for future extensibility
         if os.getenv("PAPERLESS_DBENGINE") == "mariadb":
             engine = "django.db.backends.mysql"
+            # Contrary to Postgres, Django does not natively support connection pooling for MariaDB.
+            # However, since MariaDB uses threads instead of forks, establishing connections is significantly faster
+            # compared to PostgreSQL, so the lack of pooling is not an issue
             options = {
                 "read_default_file": "/etc/mysql/my.cnf",
                 "charset": "utf8mb4",
@@ -723,6 +726,15 @@ def _parse_db_settings() -> dict:
                 "sslcert": os.getenv("PAPERLESS_DBSSLCERT", None),
                 "sslkey": os.getenv("PAPERLESS_DBSSLKEY", None),
             }
+            if int(os.getenv("PAPERLESS_DB_POOLSIZE", 0)) > 0:
+                options.update(
+                    {
+                        "pool": {
+                            "min_size": 1,
+                            "max_size": int(os.getenv("PAPERLESS_DB_POOLSIZE")),
+                        },
+                    },
+                )
 
         databases["default"]["ENGINE"] = engine
         databases["default"]["OPTIONS"].update(options)
diff --git a/uv.lock b/uv.lock
index 998649b2d..2077e1d11 100644
--- a/uv.lock
+++ b/uv.lock
@@ -1,5 +1,5 @@
 version = 1
-revision = 2
+revision = 3
 requires-python = ">=3.10"
 resolution-markers = [
     "sys_platform == 'darwin'",
@@ -340,15 +340,15 @@ wheels = [
 
 [[package]]
 name = "channels"
-version = "4.3.0"
+version = "4.3.1"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "asgiref", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "django", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/72/04/6768c7a887f9c593c4d49f99130c8aec4ea06e750bc17c306b689f6caf3b/channels-4.3.0.tar.gz", hash = "sha256:7db32c61dcd88eada1647e6c6f6ad2eb724b75d4852eeff26ad1c51ccd1a37f7", size = 26816, upload-time = "2025-07-28T13:52:50.334Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/12/a0/46450fcf9e56af18a6b0440ba49db6635419bb7bc84142c35f4143b1a66c/channels-4.3.1.tar.gz", hash = "sha256:97413ffd674542db08e16a9ef09cd86ec0113e5f8125fbd33cf0854adcf27cdb", size = 26896, upload-time = "2025-08-01T13:25:19.952Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/7c/59/0866202ee593e1b0dab0b472ebb8169e1b2b7886ad3008d193da2bbe10cb/channels-4.3.0-py3-none-any.whl", hash = "sha256:0497f3affb95e621b37d6bae1b6a5d9e8e1e1221007a2566f280091cf30ffcce", size = 31238, upload-time = "2025-07-28T13:52:49.117Z" },
+    { url = "https://files.pythonhosted.org/packages/89/1c/eae1c2a8c195760376e7f65d0bdcc3e966695d29cfbe5c54841ce5c71408/channels-4.3.1-py3-none-any.whl", hash = "sha256:b091d4b26f91d807de3e84aead7ba785314f27eaf5bac31dd51b1c956b883859", size = 31286, upload-time = "2025-08-01T13:25:18.845Z" },
 ]
 
 [[package]]
@@ -1984,6 +1984,7 @@ dependencies = [
     { name = "ocrmypdf", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "pathvalidate", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "pdf2image", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+    { name = "psycopg-pool", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "python-dateutil", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "python-dotenv", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "python-gnupg", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
@@ -2009,10 +2010,11 @@ mariadb = [
     { name = "mysqlclient", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
 ]
 postgres = [
-    { name = "psycopg", extra = ["c"], marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+    { name = "psycopg", extra = ["c", "pool"], marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "psycopg-c", version = "3.2.9", source = { registry = "https://pypi.org/simple" }, marker = "(python_full_version != '3.12.*' and platform_machine == 'aarch64' and sys_platform == 'linux') or (python_full_version != '3.12.*' and platform_machine == 'x86_64' and sys_platform == 'linux') or (platform_machine != 'aarch64' and platform_machine != 'x86_64' and sys_platform == 'linux') or sys_platform == 'darwin'" },
     { name = "psycopg-c", version = "3.2.9", source = { url = "https://github.com/paperless-ngx/builder/releases/download/psycopg-3.2.9/psycopg_c-3.2.9-cp312-cp312-linux_aarch64.whl" }, marker = "python_full_version == '3.12.*' and platform_machine == 'aarch64' and sys_platform == 'linux'" },
     { name = "psycopg-c", version = "3.2.9", source = { url = "https://github.com/paperless-ngx/builder/releases/download/psycopg-3.2.9/psycopg_c-3.2.9-cp312-cp312-linux_x86_64.whl" }, marker = "python_full_version == '3.12.*' and platform_machine == 'x86_64' and sys_platform == 'linux'" },
+    { name = "psycopg-pool", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
 ]
 webserver = [
     { name = "granian", extra = ["uvloop"], marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
@@ -2118,10 +2120,12 @@ requires-dist = [
     { name = "ocrmypdf", specifier = "~=16.10.0" },
     { name = "pathvalidate", specifier = "~=3.3.1" },
     { name = "pdf2image", specifier = "~=1.17.0" },
-    { name = "psycopg", extras = ["c"], marker = "extra == 'postgres'", specifier = "==3.2.9" },
+    { name = "psycopg", extras = ["c", "pool"], marker = "extra == 'postgres'", specifier = "==3.2.9" },
     { name = "psycopg-c", marker = "python_full_version == '3.12.*' and platform_machine == 'aarch64' and sys_platform == 'linux' and extra == 'postgres'", url = "https://github.com/paperless-ngx/builder/releases/download/psycopg-3.2.9/psycopg_c-3.2.9-cp312-cp312-linux_aarch64.whl" },
     { name = "psycopg-c", marker = "python_full_version == '3.12.*' and platform_machine == 'x86_64' and sys_platform == 'linux' and extra == 'postgres'", url = "https://github.com/paperless-ngx/builder/releases/download/psycopg-3.2.9/psycopg_c-3.2.9-cp312-cp312-linux_x86_64.whl" },
     { name = "psycopg-c", marker = "(python_full_version != '3.12.*' and platform_machine == 'aarch64' and extra == 'postgres') or (python_full_version != '3.12.*' and platform_machine == 'x86_64' and extra == 'postgres') or (platform_machine != 'aarch64' and platform_machine != 'x86_64' and extra == 'postgres') or (sys_platform != 'linux' and extra == 'postgres')", specifier = "==3.2.9" },
+    { name = "psycopg-pool" },
+    { name = "psycopg-pool", marker = "extra == 'postgres'", specifier = "==3.2.6" },
     { name = "python-dateutil", specifier = "~=2.9.0" },
     { name = "python-dotenv", specifier = "~=1.1.0" },
     { name = "python-gnupg", specifier = "~=0.5.4" },
@@ -2132,7 +2136,7 @@ requires-dist = [
     { name = "redis", extras = ["hiredis"], specifier = "~=5.2.1" },
     { name = "scikit-learn", specifier = "~=1.7.0" },
     { name = "setproctitle", specifier = "~=1.3.4" },
-    { name = "tika-client", specifier = "~=0.9.0" },
+    { name = "tika-client", specifier = "~=0.10.0" },
     { name = "tqdm", specifier = "~=4.67.1" },
     { name = "watchdog", specifier = "~=6.0" },
     { name = "whitenoise", specifier = "~=6.9" },
@@ -2472,6 +2476,9 @@ c = [
     { name = "psycopg-c", version = "3.2.9", source = { url = "https://github.com/paperless-ngx/builder/releases/download/psycopg-3.2.9/psycopg_c-3.2.9-cp312-cp312-linux_aarch64.whl" }, marker = "python_full_version == '3.12.*' and implementation_name != 'pypy' and platform_machine == 'aarch64' and sys_platform == 'linux'" },
     { name = "psycopg-c", version = "3.2.9", source = { url = "https://github.com/paperless-ngx/builder/releases/download/psycopg-3.2.9/psycopg_c-3.2.9-cp312-cp312-linux_x86_64.whl" }, marker = "python_full_version == '3.12.*' and implementation_name != 'pypy' and platform_machine == 'x86_64' and sys_platform == 'linux'" },
 ]
+pool = [
+    { name = "psycopg-pool", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+]
 
 [[package]]
 name = "psycopg-c"
@@ -2505,6 +2512,18 @@ wheels = [
     { url = "https://github.com/paperless-ngx/builder/releases/download/psycopg-3.2.9/psycopg_c-3.2.9-cp312-cp312-linux_x86_64.whl", hash = "sha256:250c357319242da102047b04c5cc78af872dbf85c2cb05abf114e1fb5f207917" },
 ]
 
+[[package]]
+name = "psycopg-pool"
+version = "3.2.6"
+source = { registry = "https://pypi.org/simple" }
+dependencies = [
+    { name = "typing-extensions", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
+]
+sdist = { url = "https://files.pythonhosted.org/packages/cf/13/1e7850bb2c69a63267c3dbf37387d3f71a00fd0e2fa55c5db14d64ba1af4/psycopg_pool-3.2.6.tar.gz", hash = "sha256:0f92a7817719517212fbfe2fd58b8c35c1850cdd2a80d36b581ba2085d9148e5", size = 29770, upload-time = "2025-02-26T12:03:47.129Z" }
+wheels = [
+    { url = "https://files.pythonhosted.org/packages/47/fd/4feb52a55c1a4bd748f2acaed1903ab54a723c47f6d0242780f4d97104d4/psycopg_pool-3.2.6-py3-none-any.whl", hash = "sha256:5887318a9f6af906d041a0b1dc1c60f8f0dda8340c2572b74e10907b51ed5da7", size = 38252, upload-time = "2025-02-26T12:03:45.073Z" },
+]
+
 [[package]]
 name = "pyasn1"
 version = "0.6.1"
@@ -2728,11 +2747,11 @@ wheels = [
 
 [[package]]
 name = "python-gnupg"
-version = "0.5.4"
+version = "0.5.5"
 source = { registry = "https://pypi.org/simple" }
-sdist = { url = "https://files.pythonhosted.org/packages/f1/3e/ba0dc69c9f4e0aeb24d93175230ef057c151790a7516012f61014918992d/python-gnupg-0.5.4.tar.gz", hash = "sha256:f2fdb5fb29615c77c2743e1cb3d9314353a6e87b10c37d238d91ae1c6feae086", size = 65705, upload-time = "2025-01-07T11:58:34.073Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/42/d0/72a14a79f26c6119b281f6ccc475a787432ef155560278e60df97ce68a86/python-gnupg-0.5.5.tar.gz", hash = "sha256:3fdcaf76f60a1b948ff8e37dc398d03cf9ce7427065d583082b92da7a4ff5a63", size = 66467, upload-time = "2025-08-04T19:26:55.778Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/7b/5b/6666ed5a0d3ce4d5444af62e373d5ba8ab253a03487c86f2f9f1078e7c31/python_gnupg-0.5.4-py2.py3-none-any.whl", hash = "sha256:40ce25cde9df29af91fe931ce9df3ce544e14a37f62b13ca878c897217b2de6c", size = 21730, upload-time = "2025-01-07T11:58:32.249Z" },
+    { url = "https://files.pythonhosted.org/packages/aa/19/c147f78cc18c8788f54d4a16a22f6c05deba85ead5672d3ddf6dcba5a5fe/python_gnupg-0.5.5-py2.py3-none-any.whl", hash = "sha256:51fa7b8831ff0914bc73d74c59b99c613de7247b91294323c39733bb85ac3fc1", size = 21916, upload-time = "2025-08-04T19:26:54.307Z" },
 ]
 
 [[package]]
@@ -3378,16 +3397,16 @@ wheels = [
 
 [[package]]
 name = "tika-client"
-version = "0.9.0"
+version = "0.10.0"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "anyio", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "httpx", marker = "sys_platform == 'darwin' or sys_platform == 'linux'" },
     { name = "typing-extensions", marker = "(python_full_version < '3.11' and sys_platform == 'darwin') or (python_full_version < '3.11' and sys_platform == 'linux')" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/94/ad/3508e42b470a037b3f5c19ca9993893d0faa30ba7ec7e6ac33db9bc3bf51/tika_client-0.9.0.tar.gz", hash = "sha256:c10bba8e40ede23c039f84ccd821fb2d290d339cc26cbd267ab9b561a1e83659", size = 2175246, upload-time = "2025-01-15T18:46:23.901Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/21/be/65bfc47e4689ecd5ead20cf47dc0084fd767b7e71e8cfabf5fddc42aae3c/tika_client-0.10.0.tar.gz", hash = "sha256:3101e8b2482ae4cb7f87be13ada970ff691bdc3404d94cd52f5e57a09c99370c", size = 2178257, upload-time = "2025-08-04T17:47:30.414Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/36/8c/90ba51e014fb548ee34dd5ed14e85ec4a205ff97b89ca393e4de321304ac/tika_client-0.9.0-py3-none-any.whl", hash = "sha256:2464e8335b5e92c276641c729e7707f1e894a2bfb51cc59abdd3bdfb532da8a0", size = 17963, upload-time = "2025-01-15T18:46:21.143Z" },
+    { url = "https://files.pythonhosted.org/packages/b1/31/002e0fa5bca67d6a19da8c294273486f6c46cbcc83d6879719a38a181461/tika_client-0.10.0-py3-none-any.whl", hash = "sha256:f5486cc884e4522575662aa295bda761bf9f101ac8d92840155b58ab8b96f6e2", size = 18237, upload-time = "2025-08-04T17:47:28.966Z" },
 ]
 
 [[package]]