TTS API comparison 2026
Cartesia Sonic 3.5 vs ElevenLabs Eleven v3 Conversational Cartesia vs ElevenLabs · text-to-speech for voice agents
Cartesia Sonic 3.5 is in the Visible delay speed tier (P95 of 320 ms or more); ElevenLabs Eleven v3 Conversational is in the no speed tier (not measured by Coval). 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 | ElevenLabs Eleven v3 Conversational |
|---|---|---|
| Speed tierCoval P95, 30 days | Visible delay | not measured |
| Median time to first audioCoval, 30-day window | 281 ms | — |
| P95 time to first audio (slow turns)Coval, 30-day window | 381 ms | — |
| Time to first audio, board chartSpeko, US East | 132 ms | 215 ms |
| Naturalness EloSpeko | 1,574 | 1,590 |
| Listener Elo (95% interval)Artificial Analysis | 1,191 ±11tie | 1,214 ±13tie |
| Numbers and dates (0 to 1)Speko | 0.18 | 0.3 |
| List price per 1M charactersCoval, Speko, Artificial Analysis | $49 to $50 | $50 |
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
- →Time to first audio, board chart: 132 ms vs 215 ms (Speko, US East)
- →List price per 1M characters: $49 to $50 vs $50 (Coval, Speko, Artificial Analysis)
Where ElevenLabs Eleven v3 Conversational is ahead
- →Naturalness Elo: 1,590 vs 1,574 (Speko)
- →Numbers and dates (0 to 1): 0.3 vs 0.18 (Speko)
Frequently asked questions
- What is the difference between Cartesia Sonic 3.5 and ElevenLabs Eleven v3 Conversational?
- Cartesia Sonic 3.5 (Cartesia) is in the Visible delay speed tier (P95 of 320 ms or more); ElevenLabs Eleven v3 Conversational (ElevenLabs) is in the no speed tier (not measured by Coval). Cartesia Sonic 3.5 leads on time to first audio, board chart and list price per 1M characters; ElevenLabs Eleven v3 Conversational leads on naturalness Elo and numbers and dates (0 to 1). They are a tie on listener Elo (95% interval). No combined score is computed.
- Which is faster for real-time voice agents, Cartesia Sonic 3.5 or ElevenLabs Eleven v3 Conversational?
- Only Cartesia Sonic 3.5 is measured by Coval (median 281 ms, P95 381 ms); ElevenLabs Eleven v3 Conversational has no speed tier, so the two cannot be compared on the same window.
- Which sounds more natural, Cartesia Sonic 3.5 or ElevenLabs Eleven v3 Conversational?
- Speko naturalness: 1,574 for Cartesia Sonic 3.5, 1,590 for ElevenLabs Eleven v3 Conversational. Artificial Analysis listener Elo: 1,191 ±11 and 1,214 ±13, a tie because the 95% intervals overlap. 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.