Practical notes on routing LLM inference with one API key — written by the team building FreeRouter.
Agent-Native already points its OpenAI engine at any OpenAI-compatible endpoint — aim OPENAI_BASE_URL at a FreeRouter key and move model and gateway choices to a dashboard rule.
September 27, 2026 · The FreeRouter Team
Pointcoin answers every prompt with two anonymous models. Turning on Parallel under Settings → MCP gave both of them web search, with no change to the app.
September 25, 2026 · The FreeRouter Team
Point Opencode at one FreeRouter key through the custom-provider dialog or opencode.json, and move model and gateway choices to a dashboard rule.
September 23, 2026 · The FreeRouter Team
What a single-gateway integration commits you to, and what changes when a FreeRouter key sits in front of the gateway you already use.
September 19, 2026 · The FreeRouter Team
Turn on Parallel web search on a chat key. Keep calling POST /v1/chat/completions. Weather, scores, and prices come from this request, not a training cutoff.
Call Jev with POST /v1/systemone on one key, branch on typed answers in code, then run a chat completion off the decision.
Learny ran on a direct OpenRouter key, migrated to FreeRouter in a few steps, then pointed the same key at Darkbloom from the dashboard.
Bring your own keys and start routing in minutes.