Route LLM chat completions through an OpenAI-compatible gateway with provider failover.
npx skills add https://github.com/antonyfmunoz/OS --skill openrouter-antonyfmunoz
Please help me install this Agent Skill. Skill: openrouter Source: https://github.com/antonyfmunoz/OS/tree/main/skills/tools/openrouter Command: npx skills add https://github.com/antonyfmunoz/OS --skill openrouter-antonyfmunoz
OpenRouter removes the hassle of managing multiple LLM provider APIs by giving you one unified endpoint that can route requests across 200+ models for the best cost, speed, and capability.
Call OpenRouter’s OpenAI-compatible chat completions endpoint using model IDs like perplexity/sonar-pro or anthropic/claude-sonnet-4-6 with your OPENROUTER_API_KEY.
High-intent search queries and answers about installing and using this skill.
You can route LLM calls across multiple providers by sending requests to an OpenAI-compatible gateway using provider-prefixed model IDs and Bearer-key authentication. This unified endpoint manages routing across 200+ models for cost and speed optimization.
To handle provider downtime and latency, use provider preference ordering and enable allow-fallbacks in your request payload. This mechanism transparently retries requests through alternative providers when the primary one is unavailable.
Yes, you can use Perplexity Sonar models for web search citations by specifying the perplexity/sonar-pro model ID. This provides web-grounded retrieval with citations, which is specifically useful for research workflows.
You do not need to heavily modify your existing OpenAI client. OpenRouter uses an OpenAI-compatible chat completions endpoint, so you only need to change the base URL, use your OPENROUTER_API_KEY, and format model IDs as provider/model-name.
Automatic failover optimizes costs and capabilities by allowing you to define a provider preference order. If a higher-priority provider fails, the gateway retries with fallbacks, balancing latency, cost, and model availability for production workflows.
Limitations include the strict requirement to use provider-prefixed model IDs (like anthropic/claude-sonnet-4-6) and the necessity of a valid OPENROUTER_API_KEY. Provider routing controls and fallbacks are optional but depend on provider availability.
glitternetworkIntegrate OpenRouter-backed LLM access into PinMe Workers with TypeScript.
scooter-lacroixSearch the web and generate cited AI-synthesized research via Perplexity AI.
serejarisCentralize OpenCode CLI provider, model, and baseURL configuration across project, global, and auth settings.
kettleofketchupDeploy an OpenAI-compatible API gateway routing requests to 100+ LLM providers.
