audio model comparison
Gemini 3.1 Flash TTS Preview vs TTS-1
Gemini 3.1 Flash TTS Preview and TTS-1 are both current audio models, from Google and OpenAI. On Artificial Analysis Speech Arena, Gemini 3.1 Flash TTS Preview scores 1214 Elo and TTS-1 1090 Elo (Artificial Analysis, measured October 13, 2026).
Data verified October 13, 2026 · Dataset 2026-10-13
Side by side
Standard-tier prices in effect today. The better value in each row is highlighted.
| Fact | Gemini 3.1 Flash TTS Preview | TTS-1 |
|---|---|---|
| Price per character | Not reported | $0.000015 |
| Price per 1M input tokens | $1 | Not reported |
| Price per 1M output tokens | $20 | Not reported |
| Artificial Analysis Speech Arena | 1214 Elo (#6 of 11) (better) | 1090 Elo (#9 of 11) |
| Provider | OpenAI | |
| Release date | April 15, 2026 | Not reported |
| Access | API only | API only |
| Context window | 8,192 tokens | Not reported |
| Max output | 16,384 tokens | Not reported |
| Input | text | text |
| Output | audio | audio |
When to choose each
Gemini 3.1 Flash TTS Preview
Best for
- Controllable text-to-speech generation
Watch out for
- This preview endpoint may change before becoming stable and has more restrictive rate limits
- Google replaced it with Gemini 3.8 Flash-Lite TTS, which is cheaper and rated higher by listeners
TTS-1
Best for
- Low-latency text-to-speech
Watch out for
- OpenAI no longer lists TTS-1 on its models overview, though its model page and $15 per 1M characters price remain
Gemini 3.1 Flash TTS Preview vs TTS-1 FAQ
Which scores higher on independent benchmarks, Gemini 3.1 Flash TTS Preview or TTS-1?
Gemini 3.1 Flash TTS Preview leads on Artificial Analysis Speech Arena, 1214 Elo to 1090 Elo. Scores are independent evaluations, not either vendor's launch figures.
Related comparisons
Production evaluation
Choosing between Gemini 3.1 Flash TTS Preview and TTS-1?
Public benchmarks narrow the shortlist. We run both on your own tasks, data and cost limits, then build the routing and review controls to run the winner in production.
- A head-to-head evaluation on your real workload
- Routing between models by cost and difficulty