Playground
The real compiler, running in this tab. Your code becomes WebAssembly here; nothing is sent anywhere.
Use a real model (off: agents use the offline mock)
Pick a provider and paste your own key to run agents against a real model. The key stays in this tab and is
sent only to that provider (or the base URL you set), straight from your browser. It is never put in a link, a
tape or this site's code. Agents with model: "mock" stay offline either way.
replaces model: "mock" with model: "claude-opus-5-5" in the editor.
Some providers refuse calls from web pages (CORS); the run then says it could not reach the API. Local
servers need to allow this page's origin: OLLAMA_ORIGINS for Ollama, --cors-style settings elsewhere. Use the
tin CLI when a provider won't answer a browser.
Press Run. Run a program with a tool fn to see the JSON Schema the compiler generates for the model. Run a program with agents to see what each one can reach. Each step of the agent loop appears here.
Agents here use model: "mock", an offline stand-in that calls the tools a prompt names, unless you add your key above.
The examples marked “recorded with Claude” replay a real session from a tape, and “scripted tape” ones replay a scripted model; both only while you leave their code unchanged.