快速结论
- GPT-Image-2.5 supports Text-to-Video, Image-to-Video, R2V, and Video Edit in one workflow.
- Use this model guide page to validate parameters with minimal request/response examples.
- Stabilize prompts here first, then move to production API integration.
关键参数
- model | string | 必填 | gpt-image-2.5-flare | - | Model identifier. Use the fixed value `gpt-image-2.5-flare`.
- prompt | string | 必填 | - | - | Main instruction describing the subject, action, scene, and style.
- aspect_ratio | string | 可选 | 16:9 | 16:9 | 9:16 | 1:1 | Video aspect ratio. Defaults to `16:9`; supports `16:9 | 9:16 | 1:1`.
- duration | integer | 可选 | 6 | 5 | 10 | 15 | 20 | Output duration in seconds. Defaults to `6`; supports `5 | 10 | 15 | 20`.
- resolution | string | 可选 | 720P | 720P | Output resolution. Defaults to `720P`; supports `720P`.
常见错误
- 400 invalid_request_error: 触发=Request payload misses required fields or has invalid field types.; 修复=Validate model, mode, and input schema before request.; 重试=Retry only after fixing payload; avoid blind retries.
- 401 authentication_error: 触发=Missing Authorization header or invalid API key.; 修复=Verify Bearer token format and API key scope.; 重试=Retry after auth fix; rotate key if repeated failures continue.
- 429 rate_limit_exceeded: 触发=Request rate, concurrency, or current quota hits upstream rate limiting.; 修复=Apply exponential backoff first, then review request rate, concurrency, and quota usage.; 重试=Use exponential backoff (e.g., 1s/2s/4s) with jitter; if it persists, reduce submission pressure.
- 500 internal_error: 触发=Transient upstream error or internal processing failure.; 修复=Capture request id and trigger retry workflow.; 重试=Retry 2-3 times with short delays; escalate if persistent.
