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

# Installation and Setup

# Installation and Setup

This guide covers how to get access to LLMTune and configure your environment for API and dashboard use. For self-hosted or internal deployments, refer to your organization’s deployment guide.

## Using the Hosted Platform

1. **Sign up** – Go to [https://llmtune.io](https://llmtune.io) and sign up or log in (email, GitHub, Google, or X).
2. **Workspace** – On first sign-in, a default workspace is created. All resources (datasets, jobs, API keys) are scoped to this workspace.
3. **API key** – In the dashboard, open **API Keys**, create a key, and copy it. Keys use the `sk_...` format. Store the key securely; it is not shown again.
4. **Base URL** – For the hosted public API, use `https://api.llmtune.io/v1` (or the URL provided by your deployment). For direct platform routes, the base is `https://llmtune.io/api`.

## Environment Variables (Integrations)

When integrating with the API from your app or scripts:

```bash theme={null}
# Required for API access
export LLMTUNE_API_KEY="sk_your_key_here"

# Optional: override base URL (e.g. for self-hosted)
export LLMTUNE_API_BASE="https://api.llmtune.io/v1"
```

Use the API key in the `Authorization` header:

```text theme={null}
Authorization: Bearer sk_your_key_here
```

## SDK / Client Usage

There is no LLMTune-specific SDK in the repository. The API is compatible with **OpenAI-style** clients:

* **Chat completions** – Use the OpenAI SDK (or similar) with `baseURL` set to `https://api.llmtune.io/v1` and `apiKey` set to your LLMTune API key.
* **Custom inference** – Use `fetch` or any HTTP client to call `POST /models/{modelId}/inference` and `POST /batch/inference` with the same auth.

See the [Inference API Guide](inference-api) for cURL, JavaScript, and Python examples.

## CLI

The repository does not include a dedicated CLI. You can drive the API from the shell using `curl` or a script. Example:

```bash theme={null}
curl -X POST "https://api.llmtune.io/v1/chat/completions" \
  -H "Authorization: Bearer $LLMTUNE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"meta-llama/Llama-3.3-70B-Instruct","messages":[{"role":"user","content":"Hello"}],"max_tokens":100}'
```

## Self-Hosted / Custom Deployments

For deployments that use the platform codebase (e.g. Next.js app + optional infra backend):

* **Platform (Next.js)** – Runs the dashboard and API routes under `/api/`. Configure environment variables (database, Stripe, inference backend, etc.) as required by the deployment.
* **Backend (optional)** – If using the separate infra/backend (e.g. for API key storage or worker coordination), set `INFRA_API_URL` / `BACKEND_API_URL` in the platform so it can forward requests.
* **Worker** – Fine-tuning execution may use a worker service (see llmtune-infra/worker). Configure `WORKER_API_URL` and related secrets in the platform.

Exact setup steps depend on your hosting environment. Refer to the [Deployment Guide](/guides/deployment) and your internal runbooks for detailed deployment instructions.

## Next Steps

* [Quickstart](quickstart) – Create a key, upload data, run a fine-tune, and call the API.
* [Authentication](../api/authentication) – API key creation and security.
* [API Overview](../api/overview) – Base URL, endpoints, and errors.