> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pinata.cloud/llms.txt
> Use this file to discover all available pages before exploring further.

# Overview

> Run LLM inference on Pinata with an OpenAI-compatible API

Pinata Inference is a hosted, OpenAI-compatible LLM endpoint. Enable it on your account, then use it two ways:

* **API** — point any OpenAI-compatible client at Pinata and call [chat completions](/inference/chat-completions) directly, no agent required.
* **Agents app** — the same hosted models show up alongside your other providers in an agent's [Models](/agents/models) list (connect the **Pinata** provider in the [Secrets Vault](/agents/secrets#connect-a-provider)).

Either way, requests run against the [models](/inference/models) Pinata hosts, and usage is metered and drawn down from your [credit balance](/inference/credits).

<Note>
  Inference is billed through credits. Before your first request, make sure your workspace has a credit balance and (optionally) [auto top-up](/inference/credits#auto-top-up) enabled so requests don't fail on an empty balance.
</Note>

## Enable inference

Turn Pinata-hosted inference on for your account with a single call, authenticated with your standard Pinata JWT:

```bash theme={null}
curl -X POST https://agents.pinata.cloud/v0/llm/enable \
  -H "Authorization: Bearer $PINATA_JWT"
```

```json Response theme={null}
{
  "success": true,
  "privateKey": "<your-inference-key>"
}
```

Enabling generates an Ed25519 key pair: the **private key is returned once** in this response and also stored encrypted in your [secrets](/agents/secrets) as `PINATA_LLM_KEY`; the public key is kept for request validation.

<Warning>
  The `privateKey` is your inference credential — it's what authenticates [chat completions](/inference/chat-completions), **not** your Pinata JWT. Save it now; it isn't returned again.
</Warning>

Once enabled, the [chat completions endpoint](/inference/chat-completions) accepts requests and usage starts drawing down your [credit balance](/inference/credits).

<Tip>
  Prefer the dashboard? Connecting the **Pinata** provider in the [Secrets Vault](/agents/secrets#connect-a-provider) does the same thing, and is also how Pinata-hosted models become selectable inside your agents.
</Tip>

### Check status

```bash theme={null}
curl https://agents.pinata.cloud/v0/llm/status \
  -H "Authorization: Bearer $PINATA_JWT"
```

```json Response theme={null}
{ "enabled": true, "createdAt": "2026-06-29T00:00:00.000Z" }
```

## Disable inference

```bash theme={null}
curl -X DELETE https://agents.pinata.cloud/v0/llm/disable \
  -H "Authorization: Bearer $PINATA_JWT"
```

```json Response theme={null}
{ "success": true }
```

Disabling revokes the inference key and removes the associated secrets. New requests to the chat completions endpoint are rejected and Pinata-hosted models are no longer selectable in agents — any running agents using the managed key are automatically flipped to the [free tier](/agents/models#free-tier-agents) fallback so they don't break. Your credit balance is untouched — disabling does not refund or expire credits — and you can re-enable at any time (which mints a **new** `privateKey`).

<Warning>
  Disabling takes effect immediately. Any in-flight requests may be cut off, and anything pointing at the endpoint will start getting errors until you re-enable.
</Warning>

## Authentication

There are two credentials, used for different routes:

| Routes                                         | Credential                                                                             |
| ---------------------------------------------- | -------------------------------------------------------------------------------------- |
| **Management** — `enable`, `disable`, `status` | Your **Pinata JWT** (from [Account → API Keys](/account-management/api-keys))          |
| **Inference** — `chat/completions`             | The **`privateKey`** returned when you enabled (stored as the `PINATA_LLM_KEY` secret) |
| **Catalog** — `models`                         | None — public                                                                          |

```http theme={null}
# Management
Authorization: Bearer <PINATA_JWT>

# Inference
Authorization: Bearer <PINATA_LLM_KEY>
```

## Next steps

<CardGroup cols={2}>
  <Card title="Chat Completions" icon="message" href="/inference/chat-completions">
    Call the OpenAI-compatible endpoint
  </Card>

  <Card title="Models" icon="microchip" href="/inference/models">
    See which models Pinata hosts
  </Card>

  <Card title="Credits" icon="coins" href="/inference/credits">
    Fund usage and set auto top-up
  </Card>
</CardGroup>