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

# Authentication

> Connect your worker to Zero Runtime and authenticate it: runtime address, auth token, API key and secret, and provider keys.

Your worker connects to the Zero Runtime runtime over a secure connection. Configure the
connection and credentials with environment variables.

## Connect to the runtime

Point the worker at your runtime address:

```bash theme={null}
export ZERORUNTIME_TARGET=us2.zeroruntime.ai:443
```

## Authenticate the worker

Get your auth token, or your API key and secret, from the
[Zero Runtime dashboard](https://app.zeroruntime.ai/). Then authenticate in one of two
ways: use a JWT auth token, or provide an API key and secret and let the SDK mint a JWT
for you.

```bash theme={null}
# JWT auth token
export ZERORUNTIME_AUTH_TOKEN=<token>
```

## Provider keys

Providers imported from `zeroruntime.inference` (Node JS: `@zeroruntime/js-sdk/inference`)
run through the [Inference Gateway](/runtime-sdk/plugins/inference/zero-runtime) and authenticate with
your `ZERORUNTIME_AUTH_TOKEN`, so they need no provider key. If you import a provider
directly from `zeroruntime.plugins`, set the key for each one your pipeline uses:

```bash theme={null}
export DEEPGRAM_API_KEY=<key>    # speech-to-text
export GOOGLE_API_KEY=<key>      # LLM
export CARTESIA_API_KEY=<key>    # text-to-speech
```

## Reference

| Variable | Purpose |
| - | - |
| `ZERORUNTIME_TARGET` | Runtime address |
| `ZERORUNTIME_AUTH_TOKEN` | JWT auth token |
| `<PROVIDER>_API_KEY` | Key for each provider you use directly from plugins (not needed for Inference Gateway providers) |


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.