ElevenLabs · Proprietary · Read 2026-10-27
ElevenLabs Flash v2.5: latency, voice quality, price and languages
- Rank
- 9 / 23
- Speed tier
- Fast enough
- P95
- 255.1 ms
- Price / 1M chars
- $40 to $50
ElevenLabs Flash v2.5 is a proprietary text-to-speech model from ElevenLabs. Coval measures a median of 191.0 ms and a P95 of 255.1 ms to first audio, which puts it in the Fast enough speed tier, rank 9 of 23 in the multilingual ranking. Artificial Analysis or Speko tested it in 9 languages.
This summary is generated from the measures below; it is not a vendor description. Measures a source does not hold are shown as n/a and are never estimated.
Speed (Coval, 30 days, US East, English)
| Speed tier | Fast enough |
| Median time to first audio | 191.0 ms |
| P90 | 232.7 ms |
| P95 | 255.1 ms |
| Mean | 197.2 ms |
| Standard deviation | 33.9 ms |
| Calls in the window | 2,798 |
| Cekura median, streamed text | 177 ms |
| Cekura median, full text | 182 ms |
Voice quality (never added together)
| Listener Elo (Artificial Analysis) | 1,076 ±11 |
| Customer-service Elo (Artificial Analysis) | 1,089 ±26 |
| Naturalness (Speko) | 1,494 |
| Naturalness estimate (Cekura, 1 to 5) | 4.33 |
Accuracy
| Reading accuracy (Cekura) | 64.4% |
| Numbers (Cekura) | 55% |
| Codes (Cekura) | 55% |
| Context (Cekura) | 75% |
| Medical terms (Cekura) | 100% |
| Hard passages judged correct (Artificial Analysis) | n/a |
| Numbers and dates (Speko score) | n/a |
| Word error rate (Coval) | 1.9% |
Price per 1 million characters
| Range across sources | $40 to $50 |
| Artificial Analysis | $50 |
| Speko | $50 |
| Coval | $40 |
| Cekura | $40 |
Coverage
| Access | Proprietary |
| Streaming | Measured |
| Coval tags | multilingual, voice-cloning |
| Rated for | Knowledge sharing, Assistants, Entertainment, Customer service, Voice agents |
| Languages tested | Arabic, Chinese, French, German, Hindi, Japanese, Portuguese, Spanish, Vietnamese |
Not collected
- 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 2026-10-27)
- https://benchmarks.cekura.ai/text-to-speech (read 2026-10-27)
- https://benchmarks-api-6wxgp27p2a-ue.a.run.app/v1/results/aggregates?benchmark=TTS&window=30d&include_series=false (read 2026-10-27)
- https://benchmarks.speko.ai/tts/ (read 2026-10-27)
- Vendor site: https://elevenlabs.io/
Keep exploring
Frequently asked questions
- How fast is ElevenLabs Flash v2.5?
- Coval measures a median time to first audio of 191.0 ms and a P95 of 255.1 ms for ElevenLabs Flash v2.5, over 30 days from the US East in English (Fast enough tier).
- How much does ElevenLabs Flash v2.5 cost?
- The benchmarks print a list price of $40 to $50 per million characters for ElevenLabs Flash v2.5.
- Which languages was ElevenLabs Flash v2.5 tested in?
- Artificial Analysis or Speko tested ElevenLabs Flash v2.5 in Arabic, Chinese, French, German, Hindi, Japanese, Portuguese, Spanish, Vietnamese.