DeepgramFish AudioAudio matchupSpeech output
Aura-2vs.Fish Audio S2.1 Pro
Which model is better?
Aura-2 and Fish Audio S2.1 Pro each lead in different tests.
Pick Aura-2 for
Pick Fish Audio S2.1 Pro 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
depends on the subtestAura-2 is clearly ahead in 0 of 14 subtests and slightly in 6, Fish Audio S2.1 Pro in 0 clearly and slightly in 6. 2 are level.
Coval TTFA leading silence p50
lower is better, as of November 6, 2026
Coval TTFA leading silence mean
lower is better, as of November 6, 2026
Coval TTFA roundtrip component maximum
lower is better, as of November 6, 2026
Show all 14 subtests
| Subtest | Aura-2 | Fish Audio S2.1 Pro | Verdict |
|---|---|---|---|
| Coval TTFA leading silence p50 (lower is better) | 133 ms | 4 ms | Fish Audio S2.1 Pro slightly ahead |
| Coval TTFA leading silence mean (lower is better) | 192 ms | 15 ms | Fish Audio S2.1 Pro slightly ahead |
| Coval TTFA roundtrip component maximum (lower is better) | 464 ms | 1,740 ms | Aura-2 slightly ahead |
| Coval TTFA leading silence maximum (lower is better) | 534 ms | 143 ms | Fish Audio S2.1 Pro slightly ahead |
| Coval TTFA roundtrip component mean (lower is better) | 115 ms | 356 ms | Aura-2 slightly ahead |
| Coval TTFA roundtrip component p50 (lower is better) | 129 ms | 333 ms | Aura-2 slightly ahead |
| Coval Time to First Audio (TTFA) maximum (lower is better) | 808 ms | 1,825 ms | Aura-2 slightly ahead |
| Coval Word Error Rate (WER) per-sample maximum (lower is better) | 68.0 % | 50.0 % | Fish Audio S2.1 Pro slightly ahead |
| Coval Time to First Audio (TTFA) p50 (lower is better) | 277 ms | 346 ms | Aura-2 slightly ahead |
| Coval Word Error Rate (WER) per-sample p95 (lower is better) | 15.2 % | 12.5 % | Fish Audio S2.1 Pro slightly ahead |
| Coval Time to First Audio (TTFA) mean (lower is better) | 307 ms | 371 ms | Aura-2 slightly ahead |
| Coval pooled corpus Word Error Rate (WER) (lower is better) | 1.9 % | 1.6 % | Fish Audio S2.1 Pro slightly ahead |
| Coval Time to First Audio (TTFA) p95 (lower is better) | 536 ms | 531 ms | level |
| Coval Word Error Rate (WER) per-sample p50 (lower is better) | 0.0 % | 0.0 % | level |
Source · latest November 6, 2026
All data side by side
Prices are verbatim from the provider, because audio models bill per minute, character or token.
| Aura-2 | Fish Audio S2.1 Pro | |
|---|---|---|
| Task | Speech output | Speech output |
| Developer | Deepgram | Fish Audio |
| Status | Active | Active |
| Released | n/a | June 23, 2026 |
| Weights | proprietary | proprietary |
| Runs locally | no | no |
| Price | n/a | Free through 2026-11-30 under fair-use rules; changes with advance notice |
| Popularity (out of 100) | 44 | 0 |