curl --request POST \
--url https://api.poyo.ai/api/generate/submit \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '
{
"model": "omni-flash",
"callback_url": "https://your-domain.com/callback",
"input": {
"prompt": "A cinematic macro shot of a glass marble rolling across a wet reflective table, realistic physics, slow camera push-in",
"resolution": "720p",
"duration": 6,
"aspect_ratio": "16:9"
}
}
'{
"code": 200,
"data": {
"task_id": "task-unified-1757165031-uyujaw3d",
"status": "not_started",
"created_time": "2026-06-03T10:30:00"
},
"message": "success"
}{
"code": 400,
"error": {
"message": "Invalid request parameters",
"type": "invalid_request_error"
}
}{
"code": 400,
"error": {
"message": "Invalid request parameters",
"type": "invalid_request_error"
}
}Google
Omni Flash
Omni Flash API 可根据文本、一张图片、恰好三张参考图片或一个输入视频，生成 4–10 秒、16:9 或 9:16、720p、1080p 或 4K 的视频。
POST
/
api
/
generate
/
submit
curl --request POST \
--url https://api.poyo.ai/api/generate/submit \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '
{
"model": "omni-flash",
"callback_url": "https://your-domain.com/callback",
"input": {
"prompt": "A cinematic macro shot of a glass marble rolling across a wet reflective table, realistic physics, slow camera push-in",
"resolution": "720p",
"duration": 6,
"aspect_ratio": "16:9"
}
}
'{
"code": 200,
"data": {
"task_id": "task-unified-1757165031-uyujaw3d",
"status": "not_started",
"created_time": "2026-06-03T10:30:00"
},
"message": "success"
}{
"code": 400,
"error": {
"message": "Invalid request parameters",
"type": "invalid_request_error"
}
}{
"code": 400,
"error": {
"message": "Invalid request parameters",
"type": "invalid_request_error"
}
}- 提交后会返回
task_id。如果提供了callback_url，当任务状态变为finished或failed时，系统会向callback_url发送 POST 请求。 - 无论是否提供
callback_url，都可以通过统一的查询任务状态接口获取结果。
omni-flash 可根据文本提示词、可选的参考图片或可选的参考视频生成短视频。
模型
- omni-flash - 文生视频、图生视频、参考图片融合和视频输入生成
输入模式
- 文生视频：只传
prompt，可选再加resolution、duration和aspect_ratio - 图生视频：传
prompt，并在image_urls中传恰好一张图片 - 参考图片融合：传
prompt，并在image_urls中传恰好三张图片 - 视频输入：传
prompt，并在video_urls中传一个视频；不要传duration
必填参数
- prompt：描述场景、运动、风格、镜头行为及其他生成指令的文本提示词
可选参数
- image_urls：图片 URL 数组。允许的数量为
0、1或3 - video_urls：视频 URL 数组。最多允许
1个视频 - resolution：
720p、1080p或4k。默认为720p - duration：
4、6、8或10秒。默认为6。提供video_urls时不要传此字段 - aspect_ratio：
16:9或9:16。16:9为横屏，9:16为竖屏。默认为16:9 - callback_url：可选的 Webhook URL，用于接收完成通知
示例
文生视频
{
"model": "omni-flash",
"callback_url": "https://your-domain.com/callback",
"input": {
"prompt": "A cinematic macro shot of a glass marble rolling across a wet reflective table, realistic physics, slow camera push-in",
"resolution": "720p",
"duration": 6,
"aspect_ratio": "16:9"
}
}
图生视频
{
"model": "omni-flash",
"input": {
"prompt": "Animate this product photo with a slow turntable motion, soft studio highlights, and subtle background parallax",
"image_urls": ["https://example.com/source-image.webp"],
"resolution": "720p",
"duration": 6,
"aspect_ratio": "16:9"
}
}
三图参考融合
{
"model": "omni-flash",
"input": {
"prompt": "Combine the character, outfit, and set references into a single fashion campaign clip with gentle camera drift",
"image_urls": [
"https://example.com/character.webp",
"https://example.com/outfit.webp",
"https://example.com/set.webp"
],
"resolution": "720p",
"duration": 8,
"aspect_ratio": "9:16"
}
}
视频输入
{
"model": "omni-flash",
"input": {
"prompt": "Use the reference motion as the timing guide and transform the final look into a clean cinematic product reveal",
"video_urls": ["https://example.com/reference-video.mp4"],
"resolution": "720p",
"aspect_ratio": "16:9"
}
}
注意事项
- 使用公开的 HTTPS 图片和视频 URL，或在提交前通过 Playground 上传文件
- 不支持在
image_urls中传两张图片 video_urls最多支持一个视频- 提供
video_urls时必须省略duration - 响应会返回
task_id；查询任务状态以获取生成的视频文件
授权
所有 API 接口都需要 Bearer Token 认证。
创建 API Key：登录 PoYo，进入 仪表盘 → API 密钥（poyo.ai/zh/dashboard/api-key），点击创建 Key。
在请求头中添加：
Authorization: Bearer YOUR_API_KEY
请求体
application/json
响应
任务提交成功
Show child attributes
Show child attributes
示例:
"success"