Prompt Latency Test
Measure true response latency for any OpenAI-compatible chat endpoint — time-to-first-token, total time and streaming throughput. Your key and prompt never leave your browser.
🔒 Runs entirely in your browser · nothing is uploaded or stored
🔒 Key safety
Your API key is used only for the request you trigger, straight from your browser to the endpoint you entered. We never see, log or store it.
⚡ TTFT vs throughput
TTFT = how fast the model starts replying (drives perceived speed). Tokens/sec = streaming rate. Both matter for chat UX.
Frequently Asked Questions
Is my API key safe?
Yes. It lives only in this page's memory and is sent solely to the endpoint you specify, directly from your browser. We have no server in the path.
Why do I get a CORS error?
Some providers block direct browser calls. Use a provider/proxy that sends CORS headers, or run the test against an endpoint you control.
Related tools