Deepgram · read on October 27, 2026
Deepgram Flux TTS
Deepgram Flux TTS is a TTS (text-to-speech) model from Deepgram. Coval measures its P95 time to first audio at 267 ms, which puts it in the Fast enough speed tier. It ranks 13 of 86 in the full ranking.
- Speed tier
- Fast enough
- P95 / median
- 267 ms / 197.7 ms
- Listener Elo
- 1,052
- Price per 1M chars
- $45
Where it ranks
Full ranking: 13 of 86. Use case ranks, by speed:
- Customer service: 9 of 70
- Voice agents: 13 of 13
- Assistants: 9 of 69
- Entertainment: 9 of 69
- Knowledge sharing: 9 of 73
The tier and the ranks are computed by this site from the benchmarks' figures, not declared by Deepgram.
Measurements and sources
Coval
| Median time to first audio (P50) | 197.7 ms |
| Time to first audio, P95 | 267 ms |
| Time to first audio, mean | 203 ms |
| Time to first audio, P90 | 246.8 ms |
| Timed requests | 2,688 |
Speko
| Naturalness score | 1,550 |
| Latency, board chart | 118 ms |
| Latency, model page sweep (median) | 106 ms |
| Numbers and dates pronunciation | 0.2 |
Artificial Analysis
| Listener Elo | 1,052 |
| Listener Elo, 95% interval | 13 |
| Listener Elo, customer service | 1,140 |
| Hard passages pronounced correctly | 72.4% |
Cekura
| Time to first audio, streamed text | 154 ms |
| Time to first audio, text sent whole | 194 ms |
| Reading accuracy | 77.4% |
| Estimated naturalness (1 to 5) | 4.16 |
Details
| Provider | Deepgram |
| Access | Proprietary |
| Streaming | A benchmark measures its time to first audio |
| Languages tested | Not collected: no benchmark here lists languages for this model |
| Measured by | Artificial Analysis, Cekura, Coval, Speko |
| Official site | deepgram.com |
Sources
- https://artificialanalysis.ai/text-to-speech (read on October 27, 2026)
- https://artificialanalysis.ai/text-to-speech/leaderboard/provider-voice (read on October 27, 2026)
- https://benchmarks.cekura.ai/text-to-speech (read on October 27, 2026)
- https://benchmarks-api-6wxgp27p2a-ue.a.run.app/v1/results/aggregates?benchmark=TTS&window=30d&include_series=false (read on October 27, 2026)
- https://benchmarks.speko.ai/tts/ (read on October 27, 2026)
Values are copied from these pages on the date shown. The match between this model and its benchmark entries is inferred, because a source writes the name differently: Cekura: “Deepgram Flux”; Speko: “Deepgram flux”. See the methodology.
Compare it with the reference model: Gradium TTS vs Deepgram Flux TTS.