About this model
OmniHuman 1.5 是一款业界领先的 lipsync 和 talking head 模型,使用입력 파라미터音轨为人像图像制作动画。它能够实现高保真、逼真的 lip-sync、自然的面部表情和流畅的头部运动,从而生成逼真的说话或演唱视频。
Cost analysis
| Provider | Cost | Notes |
|---|---|---|
| muapiapp | $0.045/sec (720p) / $0.060/sec (1080p) | 根据音频时长(5 秒至 60 秒)按秒动态计费。 |
| Fal.ai | 제공되지 않음 | 暂不可用 |
| Replicate | 제공되지 않음 | 暂不可用 |
根据音频时长(5 秒至 60 秒)按秒动态计费。
暂不可用
暂不可用
** Competitor pricing is estimated based on similar model architectures and usage tiers.
Configuration schema
| Parameter | Type | Description | Default |
|---|---|---|---|
| 프롬프트 | string | Optional prompt to guide lipsync style. | Make her sing confidently into a microphone with natural lip sync |
| 이미지 URL | string | 입력人像이미지的 URL。 | https://cdn.muapi.ai/assets/omnihuman-1-5.jpg |
| 오디오 URL | string | URL of the input audio track. | https://cdn.muapi.ai/assets/omnihuman-1-5.mp3 |
| 输出해상도 | Enum (2 options) | 출력비디오的해상도。 | 1080 |
| 快速모드 | boolean | 启用快速生成模式。 | false |
Optional prompt to guide lipsync style.
Make her sing confidently into a microphone with natural lip sync입력人像이미지的 URL。
https://cdn.muapi.ai/assets/omnihuman-1-5.jpgURL of the input audio track.
https://cdn.muapi.ai/assets/omnihuman-1-5.mp3출력비디오的해상도。
1080启用快速生成模式。
falseDeveloper documentation
Prepare Your Inputs:
Configure Parameters:
720p 或 1080p 之间选择。默认值为 1080p。pe_fast_mode)以加快迭代。-1 以随机生成。Submit Your Request:
omnihuman-1-5 endpoint。Receive and Review the Output:
Frequently asked
生成时长由입력 파라미터音轨的长度决定,最短为 5 秒,最长为 60 秒。
生成口播视频时,`image_url` 和 `audio_url` 都是必需参数。
费用根据입력 파라미터音频的时长按秒动态计算:720p 分辨率为每秒 $0.045,1080p 分辨率为每秒 $0.06。