Chat • Web Search + Memory + Metrics

Ask a question. Toggle RAG / Web Search as needed. Session memory will carry forward context.

Running Summary

Raw JSON

—

Metrics

Model —

Server total (ms) —

LLM latency (ms) —

Summary latency (ms) —

Tokens/sec —

Chars/sec —

Prompt tokens —

Completion tokens —

Total tokens —

Perf Runner

Count

Concurrency

Delay (ms)

Max Completion Tokens (optional)

—

Chat • Web Search + Memory with Metrics