character-ai/ovi/image-to-video

正式环境

https://api.302.ai

POST

/ws/api/v3/character-ai/ovi/image-to-video

InfiniteTalk 是一款基于音频驱动的对话式 AI 视频生成模型。通过单张图像和音频输入，您可以创建对话或演唱视频。我们的收费标准为每 5 秒视频生成

0.15 （ 480 p ）或

0.3（720p），并支持最长 10 分钟的视频生成。

请求参数

Header 参数

Body 参数application/json

示例

{
  "image": "https://d1q70pf5vjeyhc.cloudfront.net/media/92ecf66930134a49a5a425b9def0c266/images/1759599101378698906_HKRNJFBy.jpeg",
  "prompt": "Bright evenly lit laboratory room with metallic walls and soft white light reflections.  \nA human man in a suit stands face-to-face with a humanoid robot, both in perfect focus.  \nCamera: static medium close-up, centered framing, high exposure with clear details on both faces.  \nMood: tense, thoughtful, futuristic.  \n\n<S>We built you to understand us.<E>  \n\nA Sign\n\n<S>But sometimes I wonder if you understand us too well.<E>  \n\nThe robot tilts its head slightly, eyes glowing faint blue, voice calm and precise.  \n\n<S>Understanding is not the same as becoming.<E>  \n\n<AUDCAP>Soft ambient hum of electronics, faint mechanical servo sounds, two clear voices — human and synthetic, calm and steady<ENDAUDCAP>\n",
  "seed": -1
}

请求示例代码

Shell

JavaScript

Java

Swift

PHP

Python

HTTP

Objective-C

Ruby

OCaml

Dart

curl --location --request POST 'https://api.302.ai/ws/api/v3/character-ai/ovi/image-to-video' \
--header 'Authorization: Bearer ' \
--header 'Content-Type: application/json' \
--data-raw '{
  "image": "https://d1q70pf5vjeyhc.cloudfront.net/media/92ecf66930134a49a5a425b9def0c266/images/1759599101378698906_HKRNJFBy.jpeg",
  "prompt": "Bright evenly lit laboratory room with metallic walls and soft white light reflections.  \nA human man in a suit stands face-to-face with a humanoid robot, both in perfect focus.  \nCamera: static medium close-up, centered framing, high exposure with clear details on both faces.  \nMood: tense, thoughtful, futuristic.  \n\n<S>We built you to understand us.<E>  \n\nA Sign\n\n<S>But sometimes I wonder if you understand us too well.<E>  \n\nThe robot tilts its head slightly, eyes glowing faint blue, voice calm and precise.  \n\n<S>Understanding is not the same as becoming.<E>  \n\n<AUDCAP>Soft ambient hum of electronics, faint mechanical servo sounds, two clear voices — human and synthetic, calm and steady<ENDAUDCAP>\n",
  "seed": -1
}'

返回响应

🟢200成功

application/json

Body

示例

{
    "code": 200,
    "message": "success",
    "data": {
        "id": "83e3c2edfaf8483e83029a0ff0a758e6",
        "model": "alibaba/wan-2.5/text-to-video-fast",
        "outputs": [],
        "urls": {
            "get": "https://api.302.ai/ws/api/v3/predictions/83e3c2edfaf8483e83029a0ff0a758e6/result"
        },
        "has_nsfw_contents": [],
        "status": "created",
        "created_at": "2025-10-11T06:44:43.299Z",
        "error": "",
        "executionTime": 0,
        "timings": {
            "inference": 0
        }
    }
}

修改于 2025-10-20 08:20:47

wavespeed-ai/infinitetalk

文本生成语音(302格式)