# FlyMyAI (models)

> Generate and edit images and videos using FlyMyAI models.

Gives an agent access to FlyMyAI's image and video models and an
analysis VLM.

## What it can do

| Method | What it does |
| --- | --- |
| `generate_image` | Text → image via FlyMyAI (e.g. Flux Schnell). Returns `agent_file`; use `agent_file.public_url` or feed `agent_file_id` to `send_photo`. |
| `edit_image` | Image-to-image edit via FlyMyAI. Pass a prompt + at least one source image (HTTPS URL or base64). Returns `agent_file` with the edited result. |
| `analyze_image` | Analyze product photos / scenes with a Gemini VLM. Pass `image_urls` (HTTPS) and optional `prompt`. |
| `generate_video` | Text → video via FlyMyAI (e.g. Veo 3.1). Supports `duration_seconds`, `seed`, `aspect_ratio`, image-to-video, reference images. Returns `agent_file` with `.mp4`. |

> **Tip: Looking for the Models API?**
>
> The methods above are agent-facing **MCP actions**. For direct programmatic access to all FlyMyAI models - including schema lookup (`client.schema()`), async runs (`client.async_run()`), and the full model catalog - see the [Models API](https://docs.flymy.ai/quickstart.md) and the [Python](https://docs.flymy.ai/clients/python.md) / [JavaScript](https://docs.flymy.ai/clients/js.md) SDKs.

## How to get credentials

This MCP is built-in for FlyMyAI accounts and uses your existing account
credentials - no extra key to paste. If it shows as not configured, open
the tool and follow the on-screen prompt.

## Fields to fill in FlyMyAI

None - the MCP uses your FlyMyAI session.

## Troubleshooting

- **Rate-limited / quota exceeded** - check your plan on
  [flymy.ai](https://flymy.ai).
- **Edit refused on an uploaded image** - make sure the reference is a
  public HTTPS URL or clean base64; private URLs aren't fetched.

## Links

- [FlyMyAI dashboard](https://flymy.ai)
- [Models API docs](https://docs.flymy.ai/)

## Typed decisions

For fast classification, routing, triage, fixed-enum extraction and yes/no gates, use [TypeSafe Jev 1.13](https://docs.flymy.ai/guides/typesafe-jev). It accepts `state` and typed `questions` through its JSON Decisions endpoint and returns Choice, Noul probability or Score answers. It is not a chat model and does not generate prose, code or media. Both MCP servers expose `make_typed_decision`; agents discover `typesafe_jev.decide`.