TTS API comparison 2026
Cartesia Sonic 3.5 vs OpenAI GPT-4o mini TTS Cartesia vs OpenAI · text-to-speech for voice agents
Cartesia Sonic 3.5 is in the Visible delay speed tier (P95 of 320 ms or more); 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 | Cartesia Sonic 3.5 | OpenAI GPT-4o mini TTS |
|---|---|---|
| Speed tierCoval P95, 30 days | Visible delay | Visible delay |
| Median time to first audioCoval, 30-day window | 281 ms | 950 ms |
| P95 time to first audio (slow turns)Coval, 30-day window | 381 ms | 5,055 ms |
| Time to first audio, board chartSpeko, US East | 132 ms | 518 ms |
| Naturalness EloSpeko | 1,574 | 1,424 |
| Listener Elo (95% interval)Artificial Analysis | 1,191 ±11 | not rated |
| Numbers and dates (0 to 1)Speko | 0.18 | 0.44 |
| List price per 1M charactersCoval, Speko, Artificial Analysis | $49 to $50 | $20 |
In bold: the model ahead on that measure. “tie”: latencies within 10 ms, or Elo scores whose 95% intervals overlap.
Where Cartesia Sonic 3.5 is ahead
- →Median time to first audio: 281 ms vs 950 ms (Coval, 30-day window)
- →P95 time to first audio (slow turns): 381 ms vs 5,055 ms (Coval, 30-day window)
- →Time to first audio, board chart: 132 ms vs 518 ms (Speko, US East)
- →Naturalness Elo: 1,574 vs 1,424 (Speko)
Where OpenAI GPT-4o mini TTS is ahead
- →Numbers and dates (0 to 1): 0.44 vs 0.18 (Speko)
- →List price per 1M characters: $20 vs $49 to $50 (Coval, Speko, Artificial Analysis)
Frequently asked questions
- What is the difference between Cartesia Sonic 3.5 and OpenAI GPT-4o mini TTS?
- Cartesia Sonic 3.5 (Cartesia) is in the Visible delay speed tier (P95 of 320 ms or more); OpenAI GPT-4o mini TTS (OpenAI) is in the Visible delay speed tier (P95 of 320 ms or more). Cartesia Sonic 3.5 leads on median time to first audio, P95 time to first audio (slow turns), time to first audio, board chart and naturalness Elo; OpenAI GPT-4o mini TTS leads on numbers and dates (0 to 1) and list price per 1M characters. No combined score is computed.
- Which is faster for real-time voice agents, Cartesia Sonic 3.5 or OpenAI GPT-4o mini TTS?
- On Coval's 30-day window read on October 22, 2026, the median time to first audio is 281 ms for Cartesia Sonic 3.5 and 950 ms for OpenAI GPT-4o mini TTS (Cartesia Sonic 3.5 ahead by 669 ms); the P95, the slow turn, is 381 ms and 5,055 ms (Cartesia Sonic 3.5 ahead by 4,674 ms). Gaps under 10 ms are ties.
- Which sounds more natural, Cartesia Sonic 3.5 or OpenAI GPT-4o mini TTS?
- Speko naturalness: 1,574 for Cartesia Sonic 3.5, 1,424 for OpenAI GPT-4o mini TTS. Artificial Analysis listener Elo: 1,191 ±11 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.