LLMTest

GitHub
LLM API • LOAD TEST • CLIENT-SIDE

LLM Concurrency
Test

Fires N simultaneous requests directly from your browser and measures whether they run in parallel.

Runs entirely in your browser — no server, no relay. This only works if the API allows cross-origin (CORS) requests from a browser. If it doesn't, every request below will fail with a generic network/CORS error and there is no client-side fix for that; you'd need a server-side proxy instead. Your API key is stored only in this browser's localStorage and sent directly to the endpoint you configure — fine for ad-hoc testing, not for a key you intend to keep fully secret.

Test Configuration

Base URL is auto-resolved to the full endpoint on blur (OpenAI → /chat/completions, Anthropic → /v1/messages). A full URL you already pasted is left untouched.