Make custom app components for Modelflare: chat, image and video generation with one Modelflare Smart API Key for every model available to your account.
一个 Modelflare Smart API Key，在 Make 场景里调用你账号可用的全部模型。
Get a key / 获取 Key: https://modelflare.dev/sign-up?utm_source=make&utm_medium=plugin&utm_campaign=integrations
These files are the source of the Make app's components, in IML JSON:
| Component | Files |
|---|---|
| Base | base.iml.jsonc — base URL from the connection, bearer key, X-Modelflare-Skill attribution, error message, log sanitizing |
| Connection | connection/parameters.iml.jsonc, connection/communication.iml.jsonc — Smart API Key and base URL; verified with GET /models |
| Model RPCs | Separate text, image and video dropdowns from the key's filtered model list; legacy listModels remains for existing scenarios |
Module chat (action) |
modules/chat/* — message a text model |
Module generateImage (action) |
modules/generate-image/* — OpenAI-compatible image models |
Module createVideo / getVideo / downloadVideo (actions) |
Start a model-specific video request, check it and download the completed file |
deploy.mjs creates or updates the app through Make's SDK Apps API and uploads every component above, the icon and this README:
MAKE_API_TOKEN=... MAKE_ZONE=us2.make.com node deploy.mjsThe token needs the sdk-apps:read and sdk-apps:write scopes. The app name Make assigned on first creation is kept in make-app.json, so later runs update the same app. Making the app public (required for the invite link and for Make's review) is a one-way step in Make.
Each model's request fields are on its page at https://docs.modelflare.dev/models/. Gemini image models use generateContent and are not covered by the image module.
MIT © Havenbyte LLC