From d9a58f7b98b5b3a9d4673d7eef65ef02bcd32219 Mon Sep 17 00:00:00 2001
From: Yuneng Jiang <yuneng@berri.ai>
Date: Sat, 11 Jul 2026 13:01:09 -0700
Subject: [PATCH] docs(release_notes): add v1.90.4 release notes

---
 release_notes/v1.90.4/index.md | 54 ++++++++++++++++++++++++++++++++++
 1 file changed, 54 insertions(+)
 create mode 100644 release_notes/v1.90.4/index.md

diff --git a/release_notes/v1.90.4/index.md b/release_notes/v1.90.4/index.md
new file mode 100644
index 000000000..a4950d80a
--- /dev/null
+++ b/release_notes/v1.90.4/index.md
@@ -0,0 +1,54 @@
+---
+title: "v1.90.4 - Responses API Guardrail Coverage"
+slug: "v1-90-4"
+date: 2026-07-11T13:00:41
+authors:
+  - name: Krrish Dholakia
+    title: CEO, LiteLLM
+    url: https://www.linkedin.com/in/krish-d/
+    image_url: https://pbs.twimg.com/profile_images/1298587542745358340/DZv3Oj-h_400x400.jpg
+  - name: Ishaan Jaff
+    title: CTO, LiteLLM
+    url: https://www.linkedin.com/in/reffajnaahsi/
+    image_url: https://pbs.twimg.com/profile_images/1613813310264340481/lz54oEiB_400x400.jpg
+  - name: Yuneng Jiang
+    title: Senior Full Stack Engineer, LiteLLM
+    url: https://www.linkedin.com/in/yuneng-david-jiang-455676139/
+    image_url: https://avatars.githubusercontent.com/u/171294688?v=4
+hide_table_of_contents: false
+---
+
+## Deploy this version
+
+import Tabs from '@theme/Tabs';
+import TabItem from '@theme/TabItem';
+
+<Tabs>
+<TabItem value="docker" label="Docker">
+
+```bash
+docker run \
+-e STORE_MODEL_IN_DB=True \
+-p 4000:4000 \
+docker.litellm.ai/berriai/litellm:1.90.4
+```
+
+</TabItem>
+<TabItem value="pip" label="Pip">
+
+```bash
+pip install litellm==1.90.4
+```
+
+</TabItem>
+</Tabs>
+
+`v1.90.4` is a patch release on top of [`v1.90.3`](/release_notes/v1.90.3/v1-90-3). It restores text guardrail coverage on the Responses API (`/v1/responses`). The shared guardrail content helpers now walk the Responses `input` taxonomy (the `text`, `input_text`, and `output_text` part types, and `message`, `function_call`, and `function_call_output` items), so guardrails that build on those helpers (AIM, Lakera v2, Cato, Lasso, Repello, IBM, Azure Content Safety, and enterprise secret detection) inspect and redact request text on `/v1/responses` the same way they already do for chat completions.
+
+### What's Changed
+
+- fix(guardrails): walk Responses-API text taxonomy in shared content helpers - [PR #32542](https://github.com/BerriAI/litellm/pull/32542)
+
+## Full Changelog
+
+https://github.com/BerriAI/litellm/compare/v1.90.3...v1.90.4