PROVIDER DIRECTORY
Compare voice AI providers
The companies behind benchmarked voice AI, measured independently — no blended vendor scores and no self-reported numbers.
30 providers
Top result: Whisper Large v3 — #11 of 29 on STT TTFS 131 msDedicated inference. Shared endpoints serve many customers on the same infrastructure, while dedicated endpoints run on hardware reserved for a single customer.
Top result: Qwen3 ASR 1.7b — #1 of 29 on STT TTFS 39 msDedicated inference. Shared endpoints serve many customers on the same infrastructure, while dedicated endpoints run on hardware reserved for a single customer.
Top result: Gemini 3.5 Transcribe Live — #21 of 29 on STT TTFS 317 ms
Top result: drama3-preview — #9 of 32 on TTS TTFA 165 ms
Top result: Universal 3.6 Pro — #12 of 29 on STT TTFS 135 ms
Top result: Qwen3 ASR 1.7b — #1 of 29 on STT TTFS 39 msDedicated inference. Shared endpoints serve many customers on the same infrastructure, while dedicated endpoints run on hardware reserved for a single customer.
Top result: Flash v2.5 — #10 of 32 on TTS TTFA 191 ms
Top result: TTS Flash 2 — #4 of 32 on TTS TTFA 92 ms
Top result: Parakeet TDT 0.6B v3 — #7 of 29 on STT TTFS 97 ms
Top result: Daytona STT — #22 of 29 on STT TTFS 399 ms
Top result: Qwen3 TTS Fast — #3 of 32 on TTS TTFA 68 ms
Top result: Parakeet TDT 0.6B v3 — #7 of 29 on STT TTFS 97 ms
Top result: Simba 3.0 — #6 of 32 on TTS TTFA 111 ms
Top result: Aura 2 En — #28 of 32 on TTS TTFA 572 ms
Top result: Phantom Z 3.4 conversational — #18 of 32 on TTS TTFA 295 ms
Top result: Gemini 3.5 Transcribe Live — #21 of 29 on STT TTFS 317 ms
Top result: Voxtral Mini Transcribe Realtime 2602 — #23 of 29 on STT TTFS 410 ms
Top result: Palabra TTS v1 — #7 of 32 on TTS TTFA 119 ms
Top result: resonant-1 — #18 of 29 on STT TTFS 252 ms
Top result: Linden 1 — #20 of 29 on STT TTFS 271 ms
About this directory
“Top result” is the strongest placement any of a provider’s models holds on its category’s headline metric — TTFS, TTFA or V2V — over the last 30 days. Created models count across every host serving them; hosted models count only on that provider’s own endpoint — results are never combined into a company-level score. Individual models are compared in the model directory.
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.