# Gemini 2.5 preview TTS models shut down on 2026-11-17

- **Provider:** Google Gemini API
- **Endpoint:** `POST /*/models/gemini-2.5-*-preview-tts:*`
- **Effective:** 2026-11-17
- **Severity:** critical
- **Change:** version-sunset
- **Status:** verified

## What changed

Google shuts down gemini-2.5-flash-preview-tts and gemini-2.5-pro-preview-tts on 2026-11-17. After that the models are turned off and speech-generation calls that name them in the path fail. Calls that pass the model in a request body are not matched by this path.

## What to do

Switch to gemini-3.8-flash-tts or gemini-3.8-flash-lite-tts before 2026-11-17; the current speech-generation docs send the model in the body of POST /v1beta/interactions. Re-check voice names and how you decode the returned audio.

## Sources

- Provider notice: https://ai.google.dev/gemini-api/docs/deprecations
- https://ai.google.dev/gemini-api/docs/speech-generation

## Machine-readable

- Advisory: https://github.com/SriVarshini09/shapi-advisories/blob/main/advisories/google-gemini-2-5-preview-tts-shutdown.json
- OSV: https://shapi.dev/osv/google-gemini-2-5-preview-tts-shutdown.json