Gradium · read on October 27, 2026
Gradium TTS
Gradium TTS is a TTS (text-to-speech) model from Gradium. Coval measures its P95 time to first audio at 104.3 ms, which puts it in the Headroom speed tier. It ranks 2 of 86 in the full ranking.
- Speed tier
- Headroom
- P95 / median
- 104.3 ms / 53.7 ms
- Listener Elo
- 1,174
- Price per 1M chars
- $36 to $57.8
Where it ranks
Full ranking: 2 of 86. Use case ranks, by speed:
- Customer service: 1 of 70
- Voice agents: 2 of 13
- Assistants: 1 of 69
- Entertainment: 1 of 69
- Knowledge sharing: 1 of 73
The tier and the ranks are computed by this site from the benchmarks' figures, not declared by Gradium.
Measurements and sources
Coval
| Median time to first audio (P50) | 53.7 ms |
| Time to first audio, P95 | 104.3 ms |
| Time to first audio, mean | 64.2 ms |
| Time to first audio, P90 | 92.3 ms |
| Timed requests | 2,808 |
Speko
| Naturalness score | 1,592 |
| Latency, board chart | 55 ms |
| Latency, model page sweep (median) | 105 ms |
| Numbers and dates pronunciation | 0.27 |
Artificial Analysis
| Listener Elo | 1,174 |
| Listener Elo, 95% interval | 13 |
| Listener Elo, customer service | 1,241 |
Cekura
| Time to first audio, streamed text | -64 ms |
| Time to first audio, text sent whole | 76 ms |
| Reading accuracy | 79.5% |
| Estimated naturalness (1 to 5) | 4.51 |
Details
| Provider | Gradium |
| Access | Proprietary |
| Streaming | A benchmark measures its time to first audio |
| Features listed by Coval | multilingual, voice-cloning |
| Languages tested | Spanish, German, French, Portuguese |
| Measured by | Artificial Analysis, Cekura, Coval, Speko |
| Official site | gradium.ai |
Sources
- 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: Artificial Analysis: “Gradium TTS (Sep 2026)”; Cekura: “Gradium TTS”; Coval: “Gradium Default”; Speko: “Gradium TTS”. See the methodology.