TTS API on KeyoAPI — IndexTTS-2, CosyVoice3 & Qwen3-TTS

Looking for a TTS API you can call from the same OpenAI-style key as Whisper and chat? KeyoAPI hosts sync IndexTTS-2 plus async CosyVoice3 and Qwen3-TTS — pick latency vs batch, then drive Duix-Avatar or InfiniteTalk when you need a talking face.

Get API key Open Model Square Full pricing list Free models

TTS models on Keyo

Indicative figures for planning. Confirm live Keyo sell rates on interactive Model Square or the static pricing list.

Model ID	Shape	KeyoAPI	Notes
`IndexTTS-2`	Synchronous speech — audio in one HTTP round trip	~$1.37 / 10K characters	POST /v1/audio/speech · /pricing/IndexTTS-2
`CosyVoice3`	Async multilingual / zero-shot TTS for longer scripts	~$1.37 / 10K characters	POST /v1/async/audio/speech · /pricing/CosyVoice3
`Qwen3-TTS`	Async flagship TTS with clone-friendly voices	~$0.548 / 10K characters	POST /v1/async/audio/speech · /pricing/Qwen3-TTS

Prototype chat free, then meter TTS

Permanent $0 chat twins (deepseek-v4-flash-free, glm-5.2-free, kimi-k3-free, …) sit on the same key for script drafting. TTS rows above are metered — confirm live rates on Model Square before volume. There is no separate upload playground; try from /pricing/{modelId}.

Rates and endpoints

Teams call a TTS API for IVR, voice agents, course narration, and avatar pipelines. Keyo keeps STT (Whisper) and TTS on one prepaid balance so you are not juggling three vendor consoles.

Indicative rates above come from the live Keyo catalog — always confirm /pricing/IndexTTS-2, /pricing/CosyVoice3, and /pricing/Qwen3-TTS. Prefer IndexTTS-2 when the product needs audio inside one request; prefer CosyVoice3 or Qwen3-TTS when jobs are already async.

How to call: set base URL https://www.keyoapi.xyz/v1, authenticate with your Keyo key, pass the model id. Guides: /model/IndexTTS-2, /model/CosyVoice3, /model/Qwen3-TTS. Voice clone angle: /voice-cloning-api. Talking video: /ai-avatar-video-generator.

Create a key at /sign-up, then open the model’s /pricing/{id} page for interactive try/buy.

curl https://www.keyoapi.xyz/v1/audio/speech \
  -H "Authorization: Bearer $KEYO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"IndexTTS-2","input":"Hello from KeyoAPI TTS."}'

FAQ

Which TTS model should I start with?

Start with IndexTTS-2 for sync demos; move to CosyVoice3 or Qwen3-TTS for batch/async and clone-heavy workloads. Compare live rates on each /pricing page.

Is there a free TTS model?

Not as a permanent $0 TTS ID today. Use free chat twins to draft scripts, then meter TTS. See /free-models.

Can I pair TTS with avatars?

Yes — generate audio, then drive Duix-Avatar or InfiniteTalk on the same key. See /ai-avatar-video-generator.

Also see /compare, /free-models, and model guides under /model/.