diff --git a/doc/auto-release-notes.sh b/doc/auto-release-notes.sh
new file mode 100644
index 00000000..1a814460
--- /dev/null
+++ b/doc/auto-release-notes.sh
@@ -0,0 +1,38 @@
+#!/bin/bash
+if [ $# -eq 0 ]; then
+    echo "Usage: $0 START_DATE (format: YEAR-MM-DD) END_DATE [LABEL] [BRANCH1,BRANCH2] [LIMIT]"
+    exit 1
+fi
+
+START_DATE="$1"
+END_DATE="$2"
+
+if [ -n "$4" ]; then
+    IFS=',' read -ra BRANCHES <<< "$4"
+else
+    BRANCHES=("main")
+fi
+
+if [ -n "$5" ]; then
+    LIMIT=$5
+else
+    LIMIT=300
+fi
+
+OUTPUT=""
+for BRANCH in "${BRANCHES[@]}"; do
+    OUTPUT+=$(gh pr list --repo SpikeInterface/probeinterface --limit $LIMIT  --base "$BRANCH" --state merged --json number,title,mergedAt \
+        | jq -r --arg start_date "${START_DATE}T00:00:00Z" --arg end_date "${END_DATE}T00:00:00Z" \
+        '.[] | select(.mergedAt >= $start_date and .mergedAt <= $end_date) | "* \(.title) (#\(.number))"')
+done
+if [ -n "$OUTPUT" ]; then
+    echo ""
+    echo "$OUTPUT"
+    echo ""
+fi
+
+echo "Contributors:"
+echo ""
+gh pr list --repo SpikeInterface/probeinterface --limit 1000 --base main --state merged --json number,title,author,mergedAt \
+  | jq -r --arg start_date "${START_DATE}T00:00:00Z" --arg end_date "${END_DATE}T00:00:00Z" \
+  '[.[] | select(.mergedAt >= $start_date and .mergedAt <= $end_date and .author.login != "app/pre-commit-ci") | .author.login] | unique | .[] | "* @" + .'
diff --git a/doc/release_notes.rst b/doc/release_notes.rst
index f93343d1..01be54e9 100644
--- a/doc/release_notes.rst
+++ b/doc/release_notes.rst
@@ -6,6 +6,7 @@ Release notes
 .. toctree::
    :maxdepth: 1
 
+   releases/0.4.0.rst
    releases/0.3.2.rst
    releases/0.3.1.rst
    releases/0.3.0.rst
diff --git a/doc/releases/0.4.0.rst b/doc/releases/0.4.0.rst
new file mode 100644
index 00000000..9c95ef18
--- /dev/null
+++ b/doc/releases/0.4.0.rst
@@ -0,0 +1,59 @@
+probeinterface 0.4.0
+--------------------
+
+Jul, 6th 2026
+
+
+Features
+^^^^^^^^
+
+* Add regex validation pattern in schema version propert (#450)
+* Enhance `ProbeGroup` API and add  `ProbeGroup._global_contact_order` for interleaved contacts (#446)
+* Add Neuropixels 2.0 support to `read_spikegadgets_neuropixels` (#441)
+* Add Neuropixels saturation levels estimation from probe table/metadata (#434)
+* Auto-generate `contact_ids` when not explicitly set (#429)
+* Migrate `read_spikegadgets` to the catalogue pattern (#418)
+* Make `_parse_imro_string` self-contained by deriving IMRO field schemas from the catalogue (#417)
+* Implement copy/get_slice/get_global_contact_positions for probegroup (#416)
+* Remove legacy probe construction path for `read_imro` and support NP1110 (#410)
+
+Bug fixes
+^^^^^^^^^
+
+* Preserve planar contour on ProbeGroup slicing and clarify ambiguous `select_contacts` errors (#453)
+* Discard 'apps' folder when listing manufacturers (#451)
+* Raise on wrong-sized `device_channel_indices` instead of silently accepting (#430)
+* Preserve identity on `Probe.copy()` (#428)
+* Add contact_sides to JSON schema (#419)
+* Fix quad-base with custom names (#413)
+
+
+Deprecations
+^^^^^^^^^^^^
+
+* Rename `read_spikegadgets` to `read_spikegadgets_neuropixels` and add `has_spikegadgets_neuropixels_probes` (#440)
+* Rename `read_openephys` to `read_openephys_neuropixels` and add `has_neuropixels_probes` (#427)
+
+
+Testing
+^^^^^^^
+
+* Update CI actions (#438)
+* Add NP2020 probe to test suite (#435)
+
+Docs
+^^^
+
+* Document the Neuropixels catalogue pattern (#443)
+
+Packaging
+^^^^^^^^^
+
+* Fix typing, deprecations, and update to Python3.10 (#414)
+
+Contributors:
+
+* @alejoe91
+* @chrishalcrow
+* @h-mayorquin
+* @samuelgarcia
diff --git a/pyproject.toml b/pyproject.toml
index 51231d83..6f148c66 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [project]
 name = "probeinterface"
-version = "0.3.2"
+version = "0.4.0"
 authors = [
   { name="Samuel Garcia", email="sam.garcia.die@gmail.com" },
   { name="Alessio Buccino", email="alessiop.buccino@gmail.com" },