Evergreen Tools
|
← Back to AI Tools

OpenRouter Multi-Model API Gateway

One API key for 400+ models from every major provider, through a single OpenAI-compatible endpoint with pay-per-token billing and no monthly fee

Tool Interface

Interactive tool will be available soon

Features

  • ✓ 416+ models from Anthropic, OpenAI, Google, DeepSeek, Meta, Mistral, xAI and more
  • ✓ One OpenAI-compatible endpoint — switch models by changing base_url and the model name
  • ✓ A set of open models is available free of charge with rate limits
  • ✓ Automatic failover when a provider goes down keeps requests flowing
  • ✓ Pay-per-token with credits that never expire and no monthly subscription

How to Use

  1. Sign up for OpenRouter and create an API key
  2. Point your SDK base_url at https://openrouter.ai/api/v1
  3. Pass the model identifier in provider/model-name format
  4. Query the models endpoint to list available models and live prices

FAQ

How does OpenRouter charge?

Pay-per-token with no monthly fee. Per-token prices are set by each provider and passed through without markup; the extra costs are a 5.5% fee when buying credits and BYOK fees beyond the free threshold.

Are there free models?

Yes. OpenRouter offers a set of free models, but they carry rate limits — good for evaluation and small experiments, not as a production workhorse.

How is it different from calling providers directly?

The upside is one key, one balance and one endpoint for comparing and switching models quickly. The trade-off is an extra intermediary layer and no enterprise-grade governance, private deployment or compliance auditing. For one high-volume model, direct access can be cheaper.

Does it support streaming and tool calling?

Yes. The API stays OpenAI-compatible, so streaming, function calling and multimodal input follow each model's own capabilities — image input depends on the model you pick.

What is it best for?

Model evaluation, multi-model routing, A/B comparisons and small-to-medium production apps. Regulated teams that need evidence-backed governance and compliance should look at a dedicated LLM gateway instead.