ElevenLabs · read on October 27, 2026
ElevenLabs Flash v2.5
ElevenLabs Flash v2.5 is a TTS (text-to-speech) model from ElevenLabs. Coval measures its P95 time to first audio at 255.1 ms, which puts it in the Fast enough speed tier. It ranks 12 of 86 in the full ranking.
- Speed tier
- Fast enough
- P95 / median
- 255.1 ms / 191 ms
- Listener Elo
- 1,076
- Price per 1M chars
- $40 to $50
Where it ranks
Full ranking: 12 of 86. Use case ranks, by speed:
- Customer service: 8 of 70
- Voice agents: 12 of 13
- Assistants: 8 of 69
- Entertainment: 8 of 69
- Knowledge sharing: 8 of 73
The tier and the ranks are computed by this site from the benchmarks' figures, not declared by ElevenLabs.
Measurements and sources
Coval
| Median time to first audio (P50) | 191 ms |
| Time to first audio, P95 | 255.1 ms |
| Time to first audio, mean | 197.2 ms |
| Time to first audio, P90 | 232.7 ms |
| Timed requests | 2,798 |
Speko
| Naturalness score | 1,494 |
| Latency, board chart | 143 ms |
Artificial Analysis
| Listener Elo | 1,076 |
| Listener Elo, 95% interval | 11 |
| Listener Elo, customer service | 1,089 |
Cekura
| Time to first audio, streamed text | 177 ms |
| Time to first audio, text sent whole | 182 ms |
| Reading accuracy | 64.4% |
| Estimated naturalness (1 to 5) | 4.33 |
Details
| Provider | ElevenLabs |
| Access | Proprietary |
| Streaming | A benchmark measures its time to first audio |
| Features listed by Coval | multilingual, voice-cloning |
| Languages tested | Spanish, German, French, Portuguese, Japanese, Chinese, Arabic, Hindi, Vietnamese |
| Measured by | Artificial Analysis, Cekura, Coval, Speko |
| Official site | elevenlabs.io |
Not collected
Nothing below is estimated. Each line says which measure is missing and why.
- Numbers and dates (Speko pronunciation score): listed by Speko without this value (reading of 2026-10-27)
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: Speko: “ElevenLabs eleven_flash_v2_5”. See the methodology.
Compare it with the reference model: Gradium TTS vs ElevenLabs Flash v2.5.