22a10f3 fix: correct Azure TTS locale extraction for SSML xml:lang
📊 Changes
1 file changed (+1 additions, -1 deletions)
View changed files
📝backend/open_webui/routers/audio.py (+1 -1)
📄 Description
Summary
The Azure TTS SSML locale was being extracted using split("-")[:1], which only takes the first segment of the voice name (e.g., "en" from "en-US"). The xml:lang attribute in SSML requires a proper BCP-47 locale like "en-US", not just a bare language code.
This caused Azure TTS to either fail outright or use incorrect pronunciation/prosody rules because the SSML document had an invalid or overly broad language tag.
Fix
Changed [:1] to [:2] so the locale correctly extracts both language and region (e.g., "en-US", "zh-CN", "fr-FR").
🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.
## 📋 Pull Request Information
**Original PR:** https://github.com/open-webui/open-webui/pull/22425
**Author:** [@alvinttang](https://github.com/alvinttang)
**Created:** 3/8/2026
**Status:** ❌ Closed
**Base:** `main` ← **Head:** `fix/azure-tts-locale-truncation`
---
### 📝 Commits (1)
- [`22a10f3`](https://github.com/open-webui/open-webui/commit/22a10f31ddb3305ffc518e37f604aab80485500b) fix: correct Azure TTS locale extraction for SSML xml:lang
### 📊 Changes
**1 file changed** (+1 additions, -1 deletions)
<details>
<summary>View changed files</summary>
📝 `backend/open_webui/routers/audio.py` (+1 -1)
</details>
### 📄 Description
## Summary
The Azure TTS SSML locale was being extracted using `split("-")[:1]`, which only takes the **first** segment of the voice name (e.g., `"en"` from `"en-US"`). The `xml:lang` attribute in SSML requires a proper BCP-47 locale like `"en-US"`, not just a bare language code.
This caused Azure TTS to either fail outright or use incorrect pronunciation/prosody rules because the SSML document had an invalid or overly broad language tag.
### Fix
Changed `[:1]` to `[:2]` so the locale correctly extracts both language and region (e.g., `"en-US"`, `"zh-CN"`, `"fr-FR"`).
### Before
```python
locale = "-".join(request.app.state.config.TTS_VOICE.split("-")[:1])
# "en-US-JennyNeural" → "en" (wrong)
```
### After
```python
locale = "-".join(request.app.state.config.TTS_VOICE.split("-")[:2])
# "en-US-JennyNeural" → "en-US" (correct)
```
---
<sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
📋 Pull Request Information
Original PR: https://github.com/open-webui/open-webui/pull/22425
Author: @alvinttang
Created: 3/8/2026
Status: ❌ Closed
Base:
main← Head:fix/azure-tts-locale-truncation📝 Commits (1)
22a10f3fix: correct Azure TTS locale extraction for SSML xml:lang📊 Changes
1 file changed (+1 additions, -1 deletions)
View changed files
📝
backend/open_webui/routers/audio.py(+1 -1)📄 Description
Summary
The Azure TTS SSML locale was being extracted using
split("-")[:1], which only takes the first segment of the voice name (e.g.,"en"from"en-US"). Thexml:langattribute in SSML requires a proper BCP-47 locale like"en-US", not just a bare language code.This caused Azure TTS to either fail outright or use incorrect pronunciation/prosody rules because the SSML document had an invalid or overly broad language tag.
Fix
Changed
[:1]to[:2]so the locale correctly extracts both language and region (e.g.,"en-US","zh-CN","fr-FR").Before
After
🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.