Prompt
Run Controls
Run one profile at a time with the shared prompt to compare TTFT, prefill speed, decode speed, and total turn latency.
Last Generated Output
No runtime run yet.
AI WebGPU Lab Experiment
LLM Chat Runtime Readiness
`exp-llm-chat-runtime-shootout` now exposes a repo-specific chat runtime readiness harness. It compares deterministic WebLLM-style and Transformers.js-style profiles with the same prompt and output budget.
The profiles are synthetic on purpose. This page makes scheduling shape, TTFT, and decode behavior visible before a real runtime is wired in.
Metrics
Environment
Activity Log
Schema-Aligned Result Draft
{
"status": "pending"
}