# OpenRouter

> Unified API gateway to access hundreds of AI models across providers

OpenRouter is an API routing platform that aggregates hundreds of commercial and open-source large language models into a single OpenAI-compatible interface. It automatically routes queries across dozens of hosting providers to optimize for lowest latency, uptime, and price. Developers also get access to free tiers for various open-source models alongside prepaid pay-as-you-go options.

Website: https://openrouter.ai/

## What it is for

- Access multiple LLMs like Claude, GPT, Llama, and DeepSeek through a single API key
- Implement automatic provider fallbacks to ensure application uptime if a host goes down
- Test and prototype AI applications using free open-source model routing options
- Route prompts to the cheapest available provider for a given model to cut costs

## Features

- OpenAI-compatible unified API endpoint
- Support for 500+ models across 80+ hosting providers
- Automated provider fallbacks for high availability
- Free-tier access for selected open-source models
- Customizable data privacy policies and provider logging toggles
- Model performance and router benchmarking index
- Built-in web chat interface to test prompts directly

## Pricing

Freemium. Free access available for select open-source models; paid models billed pay-as-you-go via prepaid credits with no monthly subscription required.

## Details

- Type: Web app
- Category: [Development](https://xiry.app/topics/development.md)
- Open source: no
- Platforms: Web, API
- For: Developers and AI builders looking for a single endpoint to query diverse models without managing multiple provider accounts.
- Alternative to: Together AI, Groq, AWS Bedrock
- Tags: llm, api, routing, inference, ai models, developer tools

## Why it was saved

- [@nick_saraev, 31 July 2026](https://www.instagram.com/reel/Dbdq7RiPOKc/): Offers free API access with generous rate limits for over 20 top open-source models like Qwen, Kimi, Llama, and DeepSeek.

## Compare

- [Alternatives to Together AI](https://xiry.app/alternatives/together-ai.md)

## Related

- [AI/ML API](https://xiry.app/resources/ai-ml-api): Single API and playground for hundreds of multimodal AI models (Web app, Paid)
- [Bytez](https://xiry.app/resources/bytez): Unified serverless API to discover and run open and closed AI models (Web app, Freemium)
- [NVIDIA NIM](https://xiry.app/resources/nvidia-nim): Catalog of hosted AI inference microservices with free API endpoints. (Web app, Freemium)
- [Stripe Bill for LLM Tokens](https://xiry.app/resources/stripe-bill-for-llm-tokens): Usage-based billing infrastructure for charging end users and reselling LLM tokens. (Web app, Paid)
- [Enabling FIRE-1 in the Firecrawl Scrape API](https://xiry.app/resources/enabling-fire-1-in-the-firecrawl-scrape-api): API configuration pattern for Firecrawl's agent-driven web navigation model (Tip, Freemium)
- [Fractals](https://xiry.app/resources/fractals): Recursive task orchestrator running agent swarms in isolated git worktrees. (GitHub repo, Open source)

Updated 3 October 2026. Page: https://xiry.app/resources/openrouter

Make your own library: Xiry takes your saved reels, tweets, Reddit bookmarks and Watch Later videos, makes them searchable and pulls the useful stuff out, like this library. It's invite-only for now; join the waitlist at https://xiry.app/#waitlist.