benchr

Launch record & analysis·July 2026

Gemini 3.1 Flash TTS Preview: steerable speech with preview boundaries

Google publishes the input and output limits and supports Batch. The model card still leaves the price blank.

By the benchr team · · Changelog · Provider-published facts rechecked against the official sources on July 28, 2026

Abstract written-signal blocks unfolding into one guided ribbon of sound.
Benchr model field plate Gemini Flash TTS Text in · directed sound out
Editorial imageA BenchR editorial illustration of the transformation from a written signal to directed audio output.
Input limit8,192Text tokens
Output limit16,384Audio tokens
Batch APIYesDocumented
Launched15 Apr2026

Gemini 3.1 Flash TTS Preview launched April 15. It is a focused speech generator, not a general Gemini model that happens to speak.

Build around synthesis

The card lists text input, audio output, audio generation, and Batch support. It does not list function calling, code execution, Live API, grounding, structured output, or thinking. Keep the workflow simple: prepare text, generate audio, review the result.

Long scripts need sections

The 8,192-token input and 16,384-token output limits matter for narration and multi-speaker work. Split a long script on meaningful boundaries, keep the content map outside the model, and listen for awkward joins. One giant prompt is not a production plan.

The preview has no published rate

Google labels the model a preview, and the checked card lists no per-token price. That does not mean free. Keep cost as an unknown, watch the model and pricing pages, and get a current quote before committing a production budget.

Provider-published facts; documented gaps remain gaps
FieldVerified record
API model IDgemini-3.1-flash-tts-preview
Input / outputText / audio
Token limits8,192 input / 16,384 output
AvailabilityPreview; Batch API supported

Frequently asked

What does Gemini 3.1 Flash TTS Preview accept and return?

The official card lists text input and audio output.

What are its token limits?

Google lists 8,192 input tokens and 16,384 output tokens.

Does Google publish a per-token price for it?

No public per-token price was recorded from the checked official model page.

Changelog

  • July 28, 2026 — Published after reviewing the official provider sources and recording unreported fields as gaps.

References

  1. Official model documentation: https://ai.google.dev/gemini-api/docs/models/gemini-3.1-flash-tts-preview
  2. Official release announcement or release notes: https://ai.google.dev/gemini-api/docs/changelog
Related