KeyoAPI
InfiniteTalk API — Audio-Driven Talking Head Video
InfiniteTalk generates audio-driven talking-head video from an image or video—ideal when you need a digital human api without a motion-capture stage.
Overview
InfiniteTalk generates audio-driven talking-head video from an image or video—ideal when you need a digital human api without a motion-capture stage. Creators generate a voice with Qwen3-TTS, then animate a still spokesperson photo through KeyoAPI. Natural lip-sync and body motion make short marketing and education clips feel presentable. Async processing fits batch content calendars better than realtime WebRTC avatars. Per-request pricing around $0.342 is simple for “one clip = one charge” packaging. For long continuous streams, evaluate Duix-Avatar’s per-second meter instead. Localized ads, course intros, newsreader demos, and support explainers. Transcribe source scripts with Whisper, synthesize with TTS, animate with InfiniteTalk. Submit async jobs with model=InfiniteTalk via Keyo task APIs. Get started at /pricing/InfiniteTalk. InfiniteTalk turns an image into talking-head video via /v1/async/videos/image-to-video—ideal for support avatars and localized explainers. Image-to-talking-head is the onboarding-friendly digital human path: one portrait plus one audio track. Enforce portrait resolution minimums and reject group photos. Localize audio with Qwen3-TTS, then animate with InfiniteTalk for multilingual explainers. Store consent flags for likeness rights. For support bots, pre-render common answers; only generate live for long-tail questions. This workflow is hard to assemble across vendors—KeyoAPI packages it as one relay. Require single-subject portraits. Reject images with text overlays that break lipsync. Pre-generate FAQs; live-generate only long-tail answers to control async cost.
Quick start
Base URL: https://www.keyoapi.xyz/v1
Endpoint: POST /v1/async/videos/image-to-video
Model: model=InfiniteTalk
curl https://www.keyoapi.xyz/v1/async/videos/image-to-video \
-H "Authorization: Bearer YOUR_API_KEY" \
-F model=InfiniteTalk \
-F image=@./portrait.png \
-F audio=@./voice.wav
# then: GET /v1/task/{id}Auth uses the same Bearer API key as chat. Full notes: Keyo docs.
FAQ
What inputs are required?
Typically an image or video portrait plus driving audio. Output duration follows the audio length.
InfiniteTalk vs Duix-Avatar?
Both target digital humans; pick based on input modality and pricing unit (per request vs per second). Prototype both on KeyoAPI.
Price?
About $0.342 per request—confirm on /pricing/InfiniteTalk.