DeepgramElevenLabsAudio matchupSpeech output
Aura-2vs.Eleven v4 Turbo
Which model is better?
Eleven v4 Turbo is slightly ahead in independent tests, but the lead is not certain.
Pick Aura-2 for
Pick Eleven v4 Turbo for
Shared tests
Independent evaluators and leaderboards only, the newest shared snapshot per evaluator. With 95 % ranges a lead counts when they do not overlap; for hit rates, when it holds in a 95 % test. Without any uncertainty figure at most “slightly ahead”.
Coval
Eleven v4 Turbo slightly aheadEleven v4 Turbo is clearly ahead in 0 of 14 subtests and slightly in 10, Aura-2 in 0 clearly and slightly in 2. 2 are level.
Coval TTFA leading silence maximum
2,878 tasks, lower is better, as of November 5, 2026
Coval TTFA leading silence mean
2,878 tasks, lower is better, as of November 5, 2026
Coval Time to First Audio (TTFA) p95
2,878 tasks, lower is better, as of November 5, 2026
Show all 14 subtests
| Subtest | Aura-2 | Eleven v4 Turbo | Verdict |
|---|---|---|---|
| Coval TTFA leading silence maximum (lower is better) | 534 ms | 138 ms | Eleven v4 Turbo slightly ahead |
| Coval TTFA leading silence mean (lower is better) | 193 ms | 60 ms | Eleven v4 Turbo slightly ahead |
| Coval Time to First Audio (TTFA) p95 (lower is better) | 536 ms | 236 ms | Eleven v4 Turbo slightly ahead |
| Coval TTFA leading silence p50 (lower is better) | 134 ms | 61 ms | Eleven v4 Turbo slightly ahead |
| Coval TTFA roundtrip component maximum (lower is better) | 464 ms | 791 ms | Aura-2 slightly ahead |
| Coval Time to First Audio (TTFA) mean (lower is better) | 307 ms | 193 ms | Eleven v4 Turbo slightly ahead |
| Coval Time to First Audio (TTFA) p50 (lower is better) | 279 ms | 188 ms | Eleven v4 Turbo slightly ahead |
| Coval Word Error Rate (WER) per-sample p95 (lower is better) | 15.2 % | 12.5 % | Eleven v4 Turbo slightly ahead |
| Coval TTFA roundtrip component mean (lower is better) | 114 ms | 132 ms | Aura-2 slightly ahead |
| Coval pooled corpus Word Error Rate (WER) (lower is better) | 1.9 % | 1.7 % | Eleven v4 Turbo slightly ahead |
| Coval Word Error Rate (WER) per-sample maximum (lower is better) | 68.0 % | 62.5 % | Eleven v4 Turbo slightly ahead |
| Coval TTFA roundtrip component p50 (lower is better) | 129 ms | 124 ms | Eleven v4 Turbo slightly ahead |
| Coval Time to First Audio (TTFA) maximum (lower is better) | 808 ms | 833 ms | level |
| Coval Word Error Rate (WER) per-sample p50 (lower is better) | 0.0 % | 0.0 % | level |
Source · latest November 5, 2026
All data side by side
Prices are verbatim from the provider, because audio models bill per minute, character or token.
| Aura-2 | Eleven v4 Turbo | |
|---|---|---|
| Task | Speech output | Speech output |
| Developer | Deepgram | ElevenLabs |
| Status | Active | Active |
| Released | n/a | September 28, 2026 |
| Weights | proprietary | n/a |
| Runs locally | no | no |
| Popularity (out of 100) | 44 | 0 |