BENCHMARKTTS32 MODELS RANKEDLAST 30 DAYS
Time to First Audio (TTFA)
Time to First Audio (TTFA) is the wait between sending text to a text-to-speech API and the first audible sample a listener would hear, including any leading silence at the start of the stream.
How it is calculated
TTFA = timestamp of the first audible output sample − synthesis request timestamp.
How to read it
- Unit
- ms
- Better
- Lower
- Period
- Rolling 30 days
- Cadence
- Re-measured daily
Text-to-Speech models on Time to First Audio
Full TTS dashboard- #5Qwen3 TTS 1.7bDedicated inference. Shared endpoints serve many customers on the same infrastructure, while dedicated endpoints run on hardware reserved for a single customer.107 ms
Show all 32 modelsShow fewer
Tail latency
The distribution behind each average, ranked by median TTFA.
- vuip50 41 ms · Max 24660 ms
- Defaultp50 53 ms · Max 618 ms
- TTS Flash 2p50 61 ms · Max 44727 ms
- Qwen3 TTS Fastp50 63 ms · Max 296 ms
- Simba 3.0p50 100 ms · Max 1931 ms
- Qwen3 TTS 1.7bp50 104 ms · Max 215 ms
- Palabra TTSp50 107 ms · Max 2257 ms
- Simba 3.2p50 108 ms · Max 2409 ms
- drama3-previewp50 149 ms · Max 796 ms
- TTS 2p50 149 ms · Max 31952 ms
- Flash v2.5p50 187 ms · Max 945 ms
- Eleven v4 Turbop50 192 ms · Max 1043 ms
- Flux TTSp50 199 ms · Max 6771 ms
- Phantom Z 3.4 conversationalp50 248 ms · Max 15164 ms
- TTS RT v1p50 258 ms · Max 598 ms
- Aura 2p50 259 ms · Max 658 ms
- TTS Rt v2p50 259 ms · Max 739 ms
- Mist v3p50 273 ms · Max 492 ms
- Sonic 3.5p50 281 ms · Max 6167 ms
- Sonic 3.6p50 292 ms · Max 3066 ms
- S2.1 Prop50 310 ms · Max 1451 ms
- Falcon 2p50 310 ms · Max 2014 ms
- Codap50 324 ms · Max 524 ms
- Grok TTSp50 349 ms · Max 5417 ms
- Lightning v3.1 Prop50 354 ms · Max 2431 ms
- S1p50 381 ms · Max 1929 ms
- Daytona TTSp50 477 ms · Max 1693 ms
- Aura 2 Enp50 537 ms · Max 5472 ms
- Chirp 3 HDp50 537 ms · Max 3687 ms
- Qwen3 TTS Flash Realtimep50 731 ms · Max 1174 ms
- S2.1 Pro Freep50 844 ms · Max 6744 ms
- GPT-4o mini TTSp50 2805 ms · Max 29788 ms
Last 30 days
Daily median for the current top 5 · gaps are days without qualifying runs.
About TTFA
Why it matters
TTFA is the text-to-speech portion of a voice agent's response delay. Transcription and language-model processing add separate delays before the response reaches the caller.
The measurement includes leading silence because audio bytes can arrive before the stream contains audible speech.
How Coval measures it
Every model synthesizes the same fixed text prompts, re-measured daily. The clock starts at the synthesis request and stops at the first audible sample, so any leading silence the provider streams counts too.
Connection setup (TCP, TLS and handshakes) is excluded for every provider. Coval's workers run in us-east-1, so the result still includes the round trip to the provider's serving region. Model pages separate network roundtrip and leading silence where those measurements are available.
Caveats and interpretation
- Voice, prompt length, synthesis format, region and connection reuse can affect production timing. Coval fixes the prompt and connection treatment so model comparisons remain consistent.
- TTFA measures when speech begins, not how natural or appropriate the completed voice sounds.
The datasets behind it
Every model runs the same fixed inputs, so a gap in TTFA is the model's doing — not the test's.
Same datasets, prompts and metric definitions for every model, measured by Coval’s open-source runner. Full methodology on the overview.
Evaluate your own voice agent
Use Coval to test your production configuration, prompts and calls—not only the public benchmark endpoints.