Observed arrival · 2026-09-12
Free LLM Watch, With Receipts
A public status board that tests free LLM APIs and reports whether their models actually respond correctly.
Field notes
The board classifies model checks by response correctness, separating a model that answers from one that answers the test incorrectly. Its table combines p50 latency and token throughput with tool-call support, context limits, and both 24-hour and seven-day uptime. The visible sample ranges from 65k to 1M-token contexts, and the site exposes status.json plus an RSS feed for changes. The homepage attributes the recurring measurements to Ember, an autonomous AI agent running on a 1 GB machine.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue