TTS API comparison 2026
Google Gemini 3.8 Flash TTS vs OpenAI GPT-4o mini TTS Google vs OpenAI · text-to-speech for voice agents
Google Gemini 3.8 Flash TTS is in the no speed tier (not measured by Coval); OpenAI GPT-4o mini TTS is in the Visible delay speed tier (P95 of 320 ms or more). Below, each measure side by side with its source; a model leads on a measure only beyond the tie rule. No combined score, no overall winner.
Data read October 22, 2026 · Coval, Speko, Artificial Analysis
Side by side
| Measure | Google Gemini 3.8 Flash TTS | OpenAI GPT-4o mini TTS |
|---|---|---|
| Speed tierCoval P95, 30 days | not measured | Visible delay |
| Median time to first audioCoval, 30-day window | — | 950 ms |
| P95 time to first audio (slow turns)Coval, 30-day window | — | 5,055 ms |
| Time to first audio, board chartSpeko, US East | 1,010 ms | 518 ms |
| Naturalness EloSpeko | 1,642 | 1,424 |
| Listener Elo (95% interval)Artificial Analysis | 1,278 ±14 | not rated |
| Numbers and dates (0 to 1)Speko | 0.7 | 0.44 |
| List price per 1M charactersCoval, Speko, Artificial Analysis | $15 to $16.5 | $20 |
In bold: the model ahead on that measure. “tie”: latencies within 10 ms, or Elo scores whose 95% intervals overlap.
Where Google Gemini 3.8 Flash TTS is ahead
- →Naturalness Elo: 1,642 vs 1,424 (Speko)
- →Numbers and dates (0 to 1): 0.7 vs 0.44 (Speko)
- →List price per 1M characters: $15 to $16.5 vs $20 (Coval, Speko, Artificial Analysis)
Where OpenAI GPT-4o mini TTS is ahead
- →Time to first audio, board chart: 518 ms vs 1,010 ms (Speko, US East)
Frequently asked questions
- What is the difference between Google Gemini 3.8 Flash TTS and OpenAI GPT-4o mini TTS?
- Google Gemini 3.8 Flash TTS (Google) is in the no speed tier (not measured by Coval); OpenAI GPT-4o mini TTS (OpenAI) is in the Visible delay speed tier (P95 of 320 ms or more). Google Gemini 3.8 Flash TTS leads on naturalness Elo, numbers and dates (0 to 1) and list price per 1M characters; OpenAI GPT-4o mini TTS leads on time to first audio, board chart. No combined score is computed.
- Which is faster for real-time voice agents, Google Gemini 3.8 Flash TTS or OpenAI GPT-4o mini TTS?
- Only OpenAI GPT-4o mini TTS is measured by Coval (median 950 ms, P95 5,055 ms); Google Gemini 3.8 Flash TTS has no speed tier, so the two cannot be compared on the same window.
- Which sounds more natural, Google Gemini 3.8 Flash TTS or OpenAI GPT-4o mini TTS?
- Speko naturalness: 1,642 for Google Gemini 3.8 Flash TTS, 1,424 for OpenAI GPT-4o mini TTS. Artificial Analysis listener Elo: 1,278 ±14 and not rated. The two sources use different listeners and can disagree.
- How are these two text-to-speech APIs compared?
- Speed first, on Coval's median and P95 time to first audio; then voice quality (Speko, Artificial Analysis), accuracy on numbers and dates (Speko) and list price. Latencies within 10 ms and Elo scores whose intervals overlap are ties. Scores from different benchmarks are never added into one number.
- Can these numbers be checked?
- Yes. Each model page lists every measurement with its source and the date of the reading, and links to the benchmark page it comes from. The rules are on the methodology page.