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

# OpenRouter Provider

> Access 100+ AI models through OpenRouter — including free options

## Overview

OpenRouter gives VibeLearn access to 100+ models through a single API. Useful as a fallback when Gemini hits rate limits, or as your primary provider if you prefer a different model for concept extraction.

## Setup

### 1. Get an API Key

Sign up at [openrouter.ai/keys](https://openrouter.ai/keys).

### 2. Add to Settings

```json title="~/.vibelearn/settings.json" theme={null}
{
  "VIBELEARN_OPENROUTER_API_KEY": "sk-or-..."
}
```

## Free Models

| Model                              | Context    | Notes                         |
| ---------------------------------- | ---------- | ----------------------------- |
| `google/gemini-2.0-flash-exp:free` | 1M tokens  | Recommended free option       |
| `xiaomi/mimo-v2-flash:free`        | 32K tokens | Fast, good for short sessions |

Set a specific model:

```json theme={null}
{
  "VIBELEARN_OPENROUTER_API_KEY": "sk-or-...",
  "VIBELEARN_OPENROUTER_MODEL": "google/gemini-2.0-flash-exp:free"
}
```

## Provider Priority

If both Gemini and OpenRouter are configured, Gemini is tried first. To use OpenRouter exclusively, set only `VIBELEARN_OPENROUTER_API_KEY` (without a Gemini key).

## Troubleshooting

**No output / empty response** — Free models may be slow or at capacity. Try a different free model or use a paid model.

**`402 Payment Required`** — Model requires credits. Switch to a `:free` model or add OpenRouter credits.

**Check which provider was used:**

```bash theme={null}
grep "provider" ~/.vibelearn/logs/vibelearn-$(date +%Y-%m-%d).log | tail -5
```