# llms.txt

Machine-readable versions of these docs, for coding agents and AI tools. Reading them is free and never starts a generation.

| File | What it is |
| - | - |
| [https://vidlab.ai/vidlab-api/llms.txt](https://vidlab.ai/vidlab-api/llms.txt) | The index: every page's Markdown link with one line each. Also at `/.well-known/llms.txt`. |
| [https://vidlab.ai/vidlab-api/llms-full.txt](https://vidlab.ai/vidlab-api/llms-full.txt) | Every page of these docs as one Markdown file. |
| [https://vidlab.ai/v1/openapi.json](https://vidlab.ai/v1/openapi.json) | OpenAPI 3.1 for the `/v1` API. Also at `/vidlab-api/openapi.json`. |
| [https://vidlab.ai/vidlab-api/SKILL.md](https://vidlab.ai/vidlab-api/SKILL.md) | The [Agent skill](/vidlab-api/docs/agent-skill). |

## Markdown pages

Every docs page has a Markdown twin:

- Add `.md` to its URL: [https://vidlab.ai/vidlab-api/docs/quickstart.md](https://vidlab.ai/vidlab-api/docs/quickstart.md).
- Or request the page itself without asking for HTML, for example `curl https://vidlab.ai/vidlab-api/docs/quickstart` or `Accept: text/markdown`.
- On each section of the docs page, **Copy** puts its Markdown on your clipboard and **Markdown** opens the twin.

## Give your agent the docs

Paste this into Claude, Codex, Cursor or any agent:

```text
Read https://vidlab.ai/vidlab-api/llms.txt and the pages it links to (the exact schemas are in https://vidlab.ai/v1/openapi.json), then use the VidLab API (base URL https://vidlab.ai/v1) for what I ask. My API key is in the VIDLAB_API_KEY environment variable: read it from there and never print it. Before any paid call, get the price from POST /v1/generations/quote and tell me. Then poll GET /v1/generations/{id} until it succeeds or fails and give me the output links.
```

The prompt is not a command: paste it into your agent's chat. The agent reads the docs and writes the API calls itself.