Observed arrival · 2026-11-12
EvalLite’s three-object model for agent traces
An API for recording AI-agent runs, trace spans, and scores, with redaction defaults and usage-based event quotas.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Teams evaluating and monitoring AI agents
- Worth noticing
- The free tier stops writes at 50,000 monthly events; paid tiers accept over-cap events and record the overage.
Field notes
The page defines a Run as one agent execution, a Trace as spans such as tool calls or outputs, and a Score as a metric paired with a short reason. It says one run, span, or score counts as one event, rejects records above 64 KB without billing them, and describes API keys as visible only once, with a SHA-256 hash retained.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
$PaidCommerce or pricing visible
✦PrettyNotable craft visible
◎NicheUnusually specific use
One card from the complete issue