# Meta provider

Official OpenClaw provider plugin for the **Meta API** — an OpenAI-compatible
**Responses API** endpoint (`POST /v1/responses`).

- **Base URL:** `https://api.meta.ai/v1`
- **Auth:** `Authorization: Bearer $MODEL_API_KEY`
- **Models:** `muse-spark-1.1`, `muse-spark-1.2`, `muse-spark-1.2-contributor` (reasoning models)
  - Context window: 1,048,576 tokens (input + output share the budget)
  - Reasoning effort: `minimal | low | medium | high | xhigh` (OpenClaw default: `high`)
  - Vision: image input in `user` messages
  - Tool calling + streaming
  - Stateless encrypted reasoning replay (`store: false`)

> [!WARNING]
> Meta's [pricing documentation](https://dev.meta.ai/docs/pricing-rate-limits/) and
> [Terms of Service](https://dev.meta.ai/legal/terms-of-service) distinguish Standard
> Services from Contributor/Discounted Services:
>
> - Standard Services are the default. Meta says prompts and completions submitted to
>   Standard Services are not used to train Meta models.
> - By using Contributor/Discounted Services, you permit Meta to use Content submitted
>   to and generated by those services as described in the Terms. Under the Terms, use
>   of Discounted Services acknowledges that permission. You must not submit sensitive,
>   confidential, or personal information to the Discounted Services.
>
> Meta's [Geographic Use Policy](https://dev.meta.ai/legal/geographic-use-policy)
> governs availability. It limits API access in some jurisdictions and adds end-user
> deployment restrictions for products built with the Contributor/Discounted model;
> those additional restrictions do not apply to your own use or products built with
> Standard Services.

Meta's [model catalog](https://dev.meta.ai/docs/models) lists text, image, video,
audio, and PDF input for these models. OpenClaw's model catalog directly represents
text and image input only; the other upstream modalities are not model-manifest input
values. Meta does not publish an exact maximum-output figure for these catalog rows.

## Usage

Install the plugin and restart Gateway:

```bash
openclaw plugins install @openclaw/meta-provider
openclaw gateway restart
```

Set the API key and select the model:

```bash
export MODEL_API_KEY=<key>
```

```json5
// ~/.openclaw/openclaw.json
{
  agents: {
    defaults: {
      model: { primary: "meta/muse-spark-1.1" },
    },
  },
}
```

Or run onboarding and choose **Meta**.

## Thinking / reasoning

`--thinking <level>` and `/think <level>` map to Responses API `reasoning.effort`.
OpenClaw defaults to `high`. Meta's
[reasoning documentation](https://dev.meta.ai/docs/reasoning/) says that an
omitted effort uses a model-determined level, so `high` is an OpenClaw policy rather
than Meta's omitted-parameter default. `off` maps to `minimal` because Muse Spark does
not accept `none`.

## Docs

See `docs/providers/meta.md` for setup, onboarding, and smoke tests.

## Live test

```bash
export MODEL_API_KEY=<key>
pnpm test:live -- extensions/meta/meta.live.test.ts
```

The live suite exercises enabled Meta cases against `POST /v1/responses`.