@@ -82,22 +82,22 @@ Duplicate prevention: if a video task is already `queued` or `running` for the c
82828383## Supported providers
848485-| Provider | Default model | Text | Image ref | Video ref | API key |
86-| --------------------- | ------------------------------- | ---- | ---------------------------------------------------- | ---------------- | ---------------------------------------- |
87-| Alibaba | `wan2.6-t2v` | Yes | Yes (remote URL) | Yes (remote URL) | `MODELSTUDIO_API_KEY` |
88-| BytePlus (1.0) | `seedance-1-0-pro-250528` | Yes | Up to 2 images (I2V models only; first + last frame) | No | `BYTEPLUS_API_KEY` |
89-| BytePlus Seedance 1.5 | `seedance-1-5-pro-251215` | Yes | Up to 2 images (first + last frame via role) | No | `BYTEPLUS_API_KEY` |
90-| BytePlus Seedance 2.0 | `dreamina-seedance-2-0-260128` | Yes | Up to 9 reference images | Up to 3 videos | `BYTEPLUS_API_KEY` |
91-| ComfyUI | `workflow` | Yes | 1 image | No | `COMFY_API_KEY` or `COMFY_CLOUD_API_KEY` |
92-| fal | `fal-ai/minimax/video-01-live` | Yes | 1 image | No | `FAL_KEY` |
93-| Google | `veo-3.1-fast-generate-preview` | Yes | 1 image | 1 video | `GEMINI_API_KEY` |
94-| MiniMax | `MiniMax-Hailuo-2.3` | Yes | 1 image | No | `MINIMAX_API_KEY` or MiniMax OAuth |
95-| OpenAI | `sora-2` | Yes | 1 image | 1 video | `OPENAI_API_KEY` |
96-| Qwen | `wan2.6-t2v` | Yes | Yes (remote URL) | Yes (remote URL) | `QWEN_API_KEY` |
97-| Runway | `gen4.5` | Yes | 1 image | 1 video | `RUNWAYML_API_SECRET` |
98-| Together | `Wan-AI/Wan2.2-T2V-A14B` | Yes | 1 image | No | `TOGETHER_API_KEY` |
99-| Vydra | `veo3` | Yes | 1 image (`kling`) | No | `VYDRA_API_KEY` |
100-| xAI | `grok-imagine-video` | Yes | 1 first-frame image or up to 7 `reference_image`s | 1 video | `XAI_API_KEY` |
85+| Provider | Default model | Text | Image ref | Video ref | API key |
86+| --------------------- | ------------------------------- | ---- | ---------------------------------------------------- | ----------------------------------------------- | ---------------------------------------- |
87+| Alibaba | `wan2.6-t2v` | Yes | Yes (remote URL) | Yes (remote URL) | `MODELSTUDIO_API_KEY` |
88+| BytePlus (1.0) | `seedance-1-0-pro-250528` | Yes | Up to 2 images (I2V models only; first + last frame) | No | `BYTEPLUS_API_KEY` |
89+| BytePlus Seedance 1.5 | `seedance-1-5-pro-251215` | Yes | Up to 2 images (first + last frame via role) | No | `BYTEPLUS_API_KEY` |
90+| BytePlus Seedance 2.0 | `dreamina-seedance-2-0-260128` | Yes | Up to 9 reference images | Up to 3 videos | `BYTEPLUS_API_KEY` |
91+| ComfyUI | `workflow` | Yes | 1 image | No | `COMFY_API_KEY` or `COMFY_CLOUD_API_KEY` |
92+| fal | `fal-ai/minimax/video-01-live` | Yes | 1 image; up to 9 with Seedance reference-to-video | Up to 3 videos with Seedance reference-to-video | `FAL_KEY` |
93+| Google | `veo-3.1-fast-generate-preview` | Yes | 1 image | 1 video | `GEMINI_API_KEY` |
94+| MiniMax | `MiniMax-Hailuo-2.3` | Yes | 1 image | No | `MINIMAX_API_KEY` or MiniMax OAuth |
95+| OpenAI | `sora-2` | Yes | 1 image | 1 video | `OPENAI_API_KEY` |
96+| Qwen | `wan2.6-t2v` | Yes | Yes (remote URL) | Yes (remote URL) | `QWEN_API_KEY` |
97+| Runway | `gen4.5` | Yes | 1 image | 1 video | `RUNWAYML_API_SECRET` |
98+| Together | `Wan-AI/Wan2.2-T2V-A14B` | Yes | 1 image | No | `TOGETHER_API_KEY` |
99+| Vydra | `veo3` | Yes | 1 image (`kling`) | No | `VYDRA_API_KEY` |
100+| xAI | `grok-imagine-video` | Yes | 1 first-frame image or up to 7 `reference_image`s | 1 video | `XAI_API_KEY` |
101101102102Some providers accept additional or alternate API key env vars. See individual [provider pages](#related) for details.
103103@@ -114,7 +114,7 @@ and the shared live sweep.
114114| Alibaba | Yes | Yes | Yes | `generate`, `imageToVideo`; `videoToVideo` skipped because this provider needs remote `http(s)` video URLs |
115115| BytePlus | Yes | Yes | No | `generate`, `imageToVideo` |
116116| ComfyUI | Yes | Yes | No | Not in the shared sweep; workflow-specific coverage lives with Comfy tests |
117-| fal | Yes | Yes | No | `generate`, `imageToVideo` |
117+| fal | Yes | Yes | Yes | `generate`, `imageToVideo`; `videoToVideo` only when using Seedance reference-to-video |
118118| Google | Yes | Yes | Yes | `generate`, `imageToVideo`; shared `videoToVideo` skipped because the current buffer-backed Gemini/Veo sweep does not accept that input |
119119| MiniMax | Yes | Yes | No | `generate`, `imageToVideo` |
120120| OpenAI | Yes | Yes | Yes | `generate`, `imageToVideo`; shared `videoToVideo` skipped because this org/input path currently needs provider-side inpaint/remix access |
@@ -296,7 +296,7 @@ entries.
296296</Accordion>
297297298298<Accordion title="fal">
299-Uses a queue-backed flow for long-running jobs. Single image reference only.
299+Uses a queue-backed flow for long-running jobs. Most fal video models accept a single image reference. Seedance 2.0 reference-to-video models accept up to 9 images, 3 videos, and 3 audio references, with at most 12 total reference files.
300300</Accordion>
301301302302<Accordion title="Google (Gemini / Veo)">
@@ -349,6 +349,7 @@ capabilities: {
349349enabled: true,
350350maxVideos: 1,
351351maxInputImages: 1,
352+maxInputImagesByModel: { "provider/reference-to-video": 9 },
352353maxDurationSeconds: 5,
353354 },
354355videoToVideo: {
@@ -366,6 +367,10 @@ enough to advertise transform-mode support. Providers should declare
366367contract tests, and the shared `video_generate` tool can validate mode support
367368deterministically.
368369370+When one model in a provider has wider reference-input support than the rest,
371+use `maxInputImagesByModel`, `maxInputVideosByModel`, or
372+`maxInputAudiosByModel` instead of raising the mode-wide limit.
373+369374## Live tests
370375371376Opt-in live coverage for the shared bundled providers:
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。