Wan-2.2 Speech-to-Video 14B

Wan-S2V is a video model that generates high-quality videos from static images and audio, with realistic facial expressions, body movements, and professional camera work for film and television applications

Inference$ Commercial useFrom 15 credits per run

About Wan-2.2 Speech-to-Video 14B

audio-to-videotalking-head

Pricing

A run with the default settings costs 15 credits. The exact cost updates in the playground as you change settings, and you're only charged when you press Run. Failed runs are refunded automatically.

Inputs

prompttextarearequired

The text prompt used for video generation.

negative_prompttextareadefault: ""

Negative prompt for video generation.

num_framesintegerdefault: 80

Number of frames to generate. Must be between 40 to 120, (must be multiple of 4).

frames_per_secondintegerdefault: 16

Frames per second of the generated video. Must be between 4 to 60. When using interpolation and `adjust_fps_for_interpolation` is set to true (default true,) the final FPS will be multiplied by the number of interpolated frames plus one. For example, if the generated frames per second is 16 and the number of interpolated frames is 1, the final frames per second will be 32. If `adjust_fps_for_interpolation` is set to false, this value will be used as-is.

seedinteger

Random seed for reproducibility. If None, a random seed is chosen.

resolutionselectdefault: "480p"

Resolution of the generated video (480p, 580p, or 720p).

Options: 480p

num_inference_stepsintegerdefault: 27

Number of inference steps for sampling. Higher values give better quality but take longer.

enable_safety_checkerbooleandefault: false

If set to true, input data will be checked for safety before processing. Disabling it requires account authorization; unauthorized requests are always checked.

enable_output_safety_checkerbooleandefault: false

If set to true, output video will be checked for safety after generation.

guidance_scalenumberdefault: 3.5

Classifier-free guidance scale. Higher values give better adherence to the prompt but may decrease quality.

shiftnumberdefault: 5

Shift value for the video. Must be between 1.0 and 10.0.

video_qualityselectdefault: "high"

The quality of the output video. Higher quality means better visual quality but larger file size.

Options: low, medium, high

video_write_modeselectdefault: "balanced"

The write mode of the output video. Faster write mode means faster results but larger file size, balanced write mode is a good compromise between speed and quality, and small write mode is the slowest but produces the smallest file size.

Options: fast, balanced, small

image_urlfilerequired

URL of the input image. If the input image does not match the chosen aspect ratio, it is resized and center cropped.

audio_urlfilerequired

The URL of the audio file.

More from Wan-2.2 Speech-to-Video 14B

Wan

Wan 2.2's 14B model generates high-resolution, photorealistic images with powerful prompt understanding and fine-grained visual detail

text-to-image

Wan

Wan-2.2 video-to-video is a video model that generates high-quality videos with high visual quality and motion diversity from text prompts and source videos.

video-to-video

Wan-2.2 Text-to-Video A14B

Wan-2.2 text-to-video is a video model that generates high-quality videos with high visual quality and motion diversity from text prompts.

text to videomotion

text-to-video

Wan v2.2 A14B

fal-ai/wan/v2.2-A14B/image-to-video

image-to-video

Related audio to video models

ElevenLabs Dubbing

Generate dubbed videos or audios using ElevenLabs Dubbing feature!

dubbingaudio-to-audio

audio-to-video

LTX 2.3 Video Pro

LTX-2.3 is a high-quality, fast AI video model available in Pro and Fast variants for text-to-video, image-to-video, and audio-to-video.

stylizedtransform

audio-to-video

Ltx 2.5 Audio to Video Pro

LTX-2.5 is Lightricks' open-source audio-video model. This endpoint generates video timed to a supplied audio clip in a quality-optimized mode, for final visuals synchronized to music, dialogue, or a soundtrack.

stylizedtransform

audio-to-video

Ltx 2.5 Audio to Video Fast

LTX-2.5 is Lightricks' open-source audio-video model. This endpoint generates video timed to a supplied audio clip in a speed-optimized mode — useful for music-driven content, dialogue-led shorts, and ads keyed to a track.

stylizedtransform

audio-to-video

Longcat Single Avatar

LongCat-Video-Avatar is an audio-driven video generation model that can generates super-realistic, lip-synchronized long video generation with natural dynamics and consistent identity.

image-to-video

audio-to-video

LTX-2.3 22B

Generate video with audio from audio, text and images using LTX-2

audio-to-video

Ltx 2.3 Quality

Generate high-quality video with audio from audio, text and images using LTX-2.3

audio-to-video

LTX-2.3 22B Distilled

Generate video with audio from audio, text and images using LTX-2 Distilled

audio-to-video

PixVerse VibeMV

PixVerse VibeMV generates music videos from audio, with optional character references and lyric subtitles. It supports visual style presets, custom style references, five aspect ratios, and output at 720p or 1080p.

music-videopixverse

audio-to-video

EchoMimic V3

EchoMimic V3 generates a talking avatar model from a picture, audio and text prompt.

echomimictalking-head

audio-to-video

Avatars

Generate high-quality videos with UGC-like avatars from audio

lipsync

audio-to-video

Longcat Single Avatar

LongCat-Video-Avatar is an audio-driven video generation model that can generates super-realistic, lip-synchronized long video generation with natural dynamics and consistent identity.

audio-to-video