LogoToRouter Docs
LogoToRouter Docs
HomepageWhat is ToRouter5-minute quickstartCore concepts
Base URL & authenticationOpenAI-compatible APIAnthropic-compatible APIGemini-compatible APIStreaming responses (SSE)Model identifiers & vendor prefixesPlayground — test in the browser
Gateway

Playground — test in the browser

Send requests to any model from your browser without writing code. Useful for verifying a key, model or prompt before integrating.

The Playground is a browser-based chat UI bundled into the ToRouter portal. Pick a key, pick a model, send a message, see the response — no SDK, no curl.

Open the Playground

portal.torouter.ai/playground (sign in first).

What you can do

  • Send chat requests to any model your group allows
  • Switch between API keys to verify each one works
  • Stream the response in real time (or not)
  • Adjust temperature, max_tokens, top_p and similar parameters
  • Copy the equivalent curl command for any request you sent

Typical uses

  • "Does my key work?" — Send one message. A 200 confirms auth + group binding.
  • "Does this model work for my use case?" — Try a few prompts before wiring up code.
  • "What's the exact request?" — Use Copy as curl and paste it into the docs or a bug report.
  • "Did I break something after rotating the key?" — Compare the new key against the old behaviour.

Playground requests count against the same quota and rate limits as production requests. They show up in Usage details like any other call.

Next steps

Create a key

Get a key first.

Base URL

Move from Playground to your own code.

Usage dashboard

See every Playground request alongside production traffic.

Model identifiers & vendor prefixes

How ToRouter names models — OpenAI-style IDs by default, vendor-prefixed names for OpenRouter channels, plus the -thinking suffix.

Browse the model catalog

Discover every model ToRouter can route to, filter by modality, and open the in-console detail view.

Table of Contents

Open the PlaygroundWhat you can doTypical usesNext steps