Benchmark Almanac

Coval Text-to-Speech TTFA

fluxions/vui 42.0 tops its leaderboard Grok TTS 273.8 leads the models we track
Not in model cards

Speech models synthesize customer-service prompts. The score is median time to first audible sample in milliseconds over the published 24-hour window.

Made by
Coval
Released
April 2026
Kind
Speech recognition and translation
Level
Everyday
Scored
Checked automatically
Answers
Short answer
Tasks
All public
Last model added
April 2026
Scores refreshed
November 12th, 2026
31
Entries on its own leaderboard, 1 model we track

fluxions/vui tops the leaderboard with 42.0

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Grok TTS, at 273.8.

  1. 1 fluxions/vui 24h pooled; median TTFA (ms) 42.0
  2. 2 gradium/gradium-tts-beta-202609 24h pooled; median TTFA (ms) 48.0
  3. 3 inworld/inworld-tts-2-flash 24h pooled; median TTFA (ms) 56.0
  4. 4 deepdub/dd-etts-3.3 24h pooled; median TTFA (ms) 57.2
  5. 5 nari/qwen3-tts-fast 24h pooled; median TTFA (ms) 63.2
  6. 6 palabra/palabra-tts-v1 24h pooled; median TTFA (ms) 88.2
  7. 7 speechify/simba-3.2 24h pooled; median TTFA (ms) 99.7
  8. 8 baseten/qwen3-tts-1.7b 24h pooled; median TTFA (ms) 100.9
  9. 9 speechify/simba-3.0 24h pooled; median TTFA (ms) 107.8
  10. 10 speechify/simba-3.5 24h pooled; median TTFA (ms) 120.9
  11. 11 murf/falcon-2 24h pooled; median TTFA (ms) 142.0
  12. 12 inworld/inworld-tts-2 24h pooled; median TTFA (ms) 164.4
  13. 13 elevenlabs/eleven_v4_turbo 24h pooled; median TTFA (ms) 190.7
  14. 14 rime/coda 24h pooled; median TTFA (ms) 196.0
  15. 15 assemblyai/assemblyai-tts-1 24h pooled; median TTFA (ms) 199.6
  16. 16 elevenlabs/eleven_flash_v2_5 24h pooled; median TTFA (ms) 203.0
  17. 17 airy/airy-tts-v1 24h pooled; median TTFA (ms) 212.7
  18. 18 deepgram/flux-haley-en 24h pooled; median TTFA (ms) 215.0
  19. 19 deepgram/aura-2-thalia-en 24h pooled; median TTFA (ms) 271.5
  20. 20 guava/daytona-tts 24h pooled; median TTFA (ms) 273.5
  21. 21 Grok TTS 24h pooled; median TTFA (ms) 273.8
  22. 22 soniox/tts-rt-v2 24h pooled; median TTFA (ms) 297.6
  23. 23 soniox/tts-rt-v1 24h pooled; median TTFA (ms) 304.1
  24. 24 smallest/lightning_v3.1_pro 24h pooled; median TTFA (ms) 310.8
  25. 25 fishaudio/s2.1-pro 24h pooled; median TTFA (ms) 356.3
  26. 26 fishaudio/drama-3-preview-us-east 24h pooled; median TTFA (ms) 372.3
  27. 27 rime/mistv3 24h pooled; median TTFA (ms) 387.1
  28. 28 cloudflare/aura-2-en 24h pooled; median TTFA (ms) 495.8
  29. 29 google/chirp-3-hd 24h pooled; median TTFA (ms) 535.2
  30. 30 alibaba/qwen3-tts-flash-realtime 24h pooled; median TTFA (ms) 724.5
  31. 31 fishaudio/s2.1-pro-free 24h pooled; median TTFA (ms) 805.1

Source: Official Coval Text-to-Speech TTFA leaderboard

#EntryScoreSettingDateSource
1 fluxions/vui 42.0 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
2 gradium/gradium-tts-beta-202609 48.0 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
3 inworld/inworld-tts-2-flash 56.0 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
4 deepdub/dd-etts-3.3 57.2 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
5 nari/qwen3-tts-fast 63.2 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
6 palabra/palabra-tts-v1 88.2 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
7 speechify/simba-3.2 99.7 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
8 baseten/qwen3-tts-1.7b 100.9 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
9 speechify/simba-3.0 107.8 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
10 speechify/simba-3.5 120.9 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
11 murf/falcon-2 142.0 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
12 inworld/inworld-tts-2 164.4 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
13 elevenlabs/eleven_v4_turbo 190.7 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
14 rime/coda 196.0 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
15 assemblyai/assemblyai-tts-1 199.6 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
16 elevenlabs/eleven_flash_v2_5 203.0 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
17 airy/airy-tts-v1 212.7 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
18 deepgram/flux-haley-en 215.0 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
19 deepgram/aura-2-thalia-en 271.5 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
20 guava/daytona-tts 273.5 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
21 Grok TTS 273.8 24h pooled; median TTFA (ms) Apr 2026 Official leaderboard
22 soniox/tts-rt-v2 297.6 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
23 soniox/tts-rt-v1 304.1 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
24 smallest/lightning_v3.1_pro 310.8 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
25 fishaudio/s2.1-pro 356.3 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
26 fishaudio/drama-3-preview-us-east 372.3 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
27 rime/mistv3 387.1 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
28 cloudflare/aura-2-en 495.8 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
29 google/chirp-3-hd 535.2 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
30 alibaba/qwen3-tts-flash-realtime 724.5 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard
31 fishaudio/s2.1-pro-free 805.1 24h pooled; median TTFA (ms) Nov 2026 Official leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard
273.8 2026 273.8 2026

Source: Official Coval Text-to-Speech TTFA leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Audio and video

All 233 in Audio and video
  • Video-MMMU In use 300 expert-level lecture videos with questions testing perception, comprehension and adaptation of knowledge. Accuracy. Qwen 3.8 Max 88.7%
  • LVBench In use Hours-long video understanding with multiple-choice questions on reasoning, grounding and summarization. Gemini 4 Argon 91.7%
  • MMVU In use 3,000 expert-annotated questions on videos across 27 disciplines. Multiple-choice and open-ended accuracy. Qwen 3.8 Max 82.4%
  • MLVU In use Multi-task long video understanding (minutes to hours) with multiple-choice and generation tasks. Accuracy. Qwen 3.8 Max 90.8%
  • FLEURS In use Models transcribe read speech in 102 languages built from FLoRes-101 sentences. Model cards usually report word error rate, so lower scores are better. Phi-4-multimodal 4.00%
  • VoiceBench In use Spoken instructions with speaker, environment and content variation. Voice assistant responses scored by GPT judge or rules. Qwen3.5-Omni-Plus 93.1%