MiniMax H3 Max (Lip Sync)
MiniMax H3 Max Lip Sync makes a picture of a face speak your audio: a video as long as the audio, 480P to 2K, from $0.05 a second.
MiniMax H3 Max Lip Sync makes a talking video from one picture and one audio track. The mouth movements follow the sound, and the video is exactly as long as the audio. It takes a photo, an illustration or a character whose aspect ratio is between 0.4 and 2.5, and the video takes the supported shape nearest the picture. H3 Max is a post-trained variant of MiniMax's H3 video model, tuned for stronger prompt adherence and better aesthetics.
Use it for presenters and explainers, voiced characters and mascots, singing and music clips, and social videos. enable_transcription (on by default) transcribes the audio to guide the lips: turn it off for music, singing, laughter or noise, which the transcript may turn into words that were never spoken. Use a real person's picture or voice only with their permission: using another person's image, voice or likeness without their consent is not allowed.
Inputs: image (required: a link or an upload), audio (required: a WAV, MP3 or M4A file of 5 to 60 seconds, at most 30 at 2K, a link or an upload), resolution (480P, 768P, 1080P or 2K; 768P by default), enable_transcription and seed. Output: video, one MP4 as long as the audio, with the audio as its soundtrack, duration, its length in seconds, and seed.
Price per second of the audio you send, counted in whole seconds and rounded up: $0.05 at 480P, $0.08 at 768P, $0.16 at 1080P and $0.32 at 2K, and 1.2 times as much for a video over 15 seconds (the whole video). A 5-second clip at 768P costs $0.40, a 12.3-second clip (13 seconds) $1.04 and a 30-second clip $2.88. The price follows the length of the audio you send, read before the run.
Here one request takes audio of up to 60 seconds (30 at 2K): split a longer track into parts. It does not write or voice the speech (bring a recording, or make one with a text-to-speech model), take a prompt, or re-sync the lips of an existing video (use Sync Lipsync 3 (Video)).
- Image to Video
Price: From $0.05 per second of the audio you send, rounded up, depending on resolution. A 10-second clip costs $0.80 with the default settings.
Run MiniMax H3 Max (Lip Sync) via API
curl -X POST https://api.flymy.ai/api/v1/flymyai/minimax-h3-max-lip-sync/predict \
-H "X-API-KEY: $FLYMYAI_API_KEY" \
-H "Content-Type: application/json" \
-d '{"image":"https://storage.googleapis.com/open-inference-results/agent-files/99/ingests/954c188c570e5d1ab609546c0edfe481/14ba39b00b65d9bfe83550ebc5c1ccb0692a0658334c9a129e1e9ca544037ef5/omnihuman2-demo-presenter.jpg","audio":"https://v3.fal.media/files/rabbit/Ql3ade3wEKlZXRQLRbhxm_tts.mp3","resolution":"480P"}'API documentation · Create a free account · Pricing