Pick or switch between vendors' video models through one key
One company's docs show a full fit.
What counts as done
The agent reads each candidate model's schema and price in one call per model, runs the same prompt on two or more models with one credential, and the platform exposes the model's current parameters.
How agents ask for it: which model is best for faces
, try this prompt on Kling and Veo
, use the cheapest model that does sound
, use the new Seedance
Does the job 1
The docs show it meets the test above, and nothing we read says otherwise.
Does the job, with a limit 6
It does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.
| Listing | Why |
|---|---|
| fal video models BB Features & Labels, Inc. | One key reaches many models, and per-model llms.txt and the pricing API give schema and price, but billing units differ (per second, per audio second, per 1,000 tokens) and duration is typed differently per model. source |
| Higgsfield API B Higgsfield Inc. | One key reaches Kling, Seedance, Hailuo, HappyHorse, Grok and Higgsfield models with an estimate per model, but each model's schema is on its own docs page rather than one call, and web-app motion presets have no API equivalent per a first-hand report. source |
| Magic Hour MCP D Magic Hour AI, Inc. | One key and one tool shape reach about 25 video models, with each model's durations, resolutions and audio rules in the tools/list schema, but per-model credit prices aren't returned before a call. source |
| Popcraft MCP F Popcraft | popcraft_models_explore returns capabilities, reference limits, durations and credit costs per model for Seedance, Veo, Kling, Wan and Hailuo under one sign-in, but published tool inventories disagree and the live schemas weren't read. source |
| Replicate video models B Replicate, LLC | One token and one predictions shape reach Veo, Kling, Seedance, Wan and Runway models, and each model has an llms.txt schema, but the price isn't in that file and input names differ per model. source |
| Runway API C Runway AI, Inc. | One key reaches Gen-4.5, Veo, Kling, Seedance, Wan and others, and the router compares by cost or quality, but request bodies differ per model and prices sit on a separate page rather than a per-model call. source |
Doesn't do this job 12
Listings in the category that do something else, with the reason.
| Listing | Why |
|---|---|
| Alibaba Wan (Model Studio) B Alibaba Cloud (Singapore) Private Limited | Alibaba's own Wan models only; no cross-vendor routing. source |
| Clueso F Desklamp, Inc. | No model catalogue; Clueso is an editor, not a model platform. source |
| framefields C Gatewai | The library is a renderer; the gateway's model catalogue isn't readable without a browser. source |
| Google Veo BB Google LLC | One vendor's model family; it does not run other vendors' models under one credential. source |
| HyperFrames E HeyGen | A renderer, not a model platform; no model catalogue or per-model price. source |
| Kling AI API F Kling AI Pte. Ltd. | Kuaishou's own Kling models only; no cross-vendor routing. source |
| Luma AI API D Luma AI, Inc. | Ray 3.2 is the only public video model on the API; it does not route across vendors. source |
| MiniMax Video API B Nanonoble Pte. Ltd. | A single vendor's H3 family; it does not run other vendors' models under one credential. source |
| PixVerse API E MOTIVAI PRIVATE LIMITED | PixVerse's own models only; no cross-vendor routing. source |
| Seedance on BytePlus ModelArk C BytePlus Pte. Ltd. | ModelArk's video endpoint serves ByteDance's Seedance family only; it does not route across vendors. source |
| sync. labs MCP server D Synchronicity Labs | sync.'s own lip-sync models only (models_list); no cross-vendor video generation. source |
| Vidu API E ShengShu AI HK Limited | One vendor's Vidu models only; no cross-vendor routing. source |
1 listings in the category were added after this reading and aren't marked yet.
What stops agents finishing
- platform lags the vendor so new parameters or presets are missing
- billing unit differs per model so prices cannot be compared before the call
- new model reachable only in a consumer app
Around it
Before: video-generation.text-to-clip. After: video-generation.automated-content-pipeline.
Where this comes from
Weight 12 of 100 in Video generation: 6 of 37 accounts; aggregator users on fal and Higgsfield, cost-in-response requests, and access questions for new models
The same as data: /api/v1/use-cases/video-generation.model-routing.json, or the MCP tool get_use_case with id video-generation.model-routing. Other jobs in this category: /api/v1/niches?category=video-generation.
For companies
Do agents find, use and choose your tools?
An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.