Cloudflare's Clairic extraction model returns confidence scores and spans, no preset categories
altryne · x · 2026-11-18
Cloudflare launched Clairic, a generalized extraction model: unlike classifiers that need predefined categories, it can extract from never-seen content (e.g., unread emails) and returns a confidence score plus a character span. Michelle Chen highlights the broader industry trend — LiquidAI shipping extraction models and OpenRouter supporting extraction endpoints — arguing everyone wins as agents get more reliable and autonomous. Covered in an interview on the Thursdai podcast.
Related event: Cloudflare Releases Universal Extraction Model Clairic(3 posts)→
More from coding & agent
- Agent security testing thread: capture evidence before deleting the dependency graph — krishnan · 2026-11-18
- Throwaway analysis environments: agents should leave reproducible evidence, no running servers — krishnan · 2026-11-18
- New plug-ins appear in Google's Antigravity IDE — DynamicWebPaige · 2026-11-18
- Execution is cheap now, so be bolder with your UI design explorations — jh3yy · 2026-11-18
- Salesforce brings rich interactive experiences to Slackbot via Headless Experience Layer — msrivastav13 · 2026-11-18
- Agent Manager: a weekly agent that audits all your other agents, with sample prompt — nlarusstone · 2026-11-18