gemini-3.1-flash-tts-preview appears to fail for Hebrew TTS prompts. The API returns HTTP 200, but the response content is empty and finishReason is OTHER, so no usable audio is returned.
Forum report with full details:
https://discuss.ai.google.dev/t/bug-gemini-3-1-flash-tts-preview-returns-http-200-with-empty-content-and-finishreason-other-for-hebrew-tts-prompt/144297
Steps to reproduce:
Call the Gemini API with model gemini-3.1-flash-tts-preview.
Use a Hebrew TTS prompt.
Request audio/TTS output.
Observe that the request succeeds with HTTP 200, but the response contains no usable audio/content.
Expected behavior:
The model should return valid audio content for the Hebrew TTS prompt, or return a clear error if Hebrew TTS is not supported.
Actual behavior:
The API returns HTTP 200, but the response content is empty and the finish reason is OTHER.
Environment:
Product: Gemini API / TTS
Model: gemini-3.1-flash-tts-preview
Language: Hebrew
Issue type: TTS generation returns empty content despite successful HTTP response
Question:
Is this a known issue with Hebrew TTS or with bracketed performance tags in TTS prompts? Is there a workaround or a better place to track this bug?
I am calling gemini-3.1-flash-tts-preview for Hebrew TTS generation. The API returns HTTP 200 and reports generated AUDIO tokens, but the candidate content is empty and the finishReason is OTHER. This makes the request look successful at the HTTP level, but there is no usable audio/content in the response. Model gemini-3.1-flash-tts-previe...