What OpenRouter solves
Managing multiple LLMLLMLarge Language Model — a neural network trained on vast amounts of text data that can understand and generate human-quality language across a wide range of tasks.Learn more → providers means multiple API keys, multiple SDKs, different rate limits, and different pricing. OpenRouter unifies all of them behind a single OpenAI-compatible API endpoint. One key, one SDK, access to GPTGPTGenerative Pre-trained Transformer — the model architecture and family name behind OpenAI's most famous models, from GPT-2 to GPT-5.Learn more →-4o, Claude, Gemini, Llama, Mistral, DeepSeek, and hundreds more.
OpenRouter also adds a routing layer: you can set fallbacks so if Claude is unavailable, requests automatically route to GPT-4o; or use cost routing to automatically select the cheapest model that meets your quality requirement.
Set up OpenRouter
Create an account at openrouter.ai. Add credits (minimum $5) and generate an API key. Set it as `OPENROUTER_API_KEY` in your environment.
OpenRouter is fully OpenAI API-compatible. Use the OpenAI SDK with a custom base URL: Python: `client = OpenAI(base_url='https://openrouter.ai/api/v1', api_key=os.environ['OPENROUTER_API_KEY'])`. Node: `const client = new OpenAI({baseURL: 'https://openrouter.ai/api/v1', apiKey: process.env.OPENROUTER_API_KEY})`.
Select and compare models
Model IDs on OpenRouter follow the format `provider/model-name`. Examples: `anthropic/claude-sonnet-4-5`, `openai/gpt-4o`, `meta-llama/llama-3.3-70b-instruct`, `deepseek/deepseek-r1`, `google/gemini-2.5-flash`.
The OpenRouter model list at openrouter.ai/models shows pricing, context length, and benchmarks for every available model. Filter by price, context windowContext WindowThe maximum amount of text (measured in tokens) that a model can read and reason over in a single interaction, including your prompt, any documents, and previous conversation history.Learn more →, or provider to find the right model for your use case.
Cost optimisation with auto-routing
Use `openai/auto` as the model to let OpenRouter automatically select the cheapest model that can handle your request. Add a `route` header to specify preferences: `{'HTTP-Referer': 'your-app', 'X-Title': 'your-app', 'route': 'cheapest'}`.
For tasks where quality varies a lot by model, use the model ranking feature: pass a list of models in order of preference with a fallback option. OpenRouter tries the first model; if it is unavailable, it tries the next.