Cloudflare's Clairic: a System One extraction model returning typed values with confidence scores
omarsar0 · x · 2026-10-27
- DAIR.AI's omarsar0 published a tutorial on Clairic, an extraction model Cloudflare introduced last week.
- Clairic is a "System One" model: it reads unstructured text and images in one pass and returns each requested field (order number, total, date, etc.) in the requested type with a confidence score — no text generation. Available in preview on Workers AI as @cf/cloudflare/clairic.
- It complements TypeSafe's Jev, the best-known System One model: Jev picks from predefined options (approve/review/reject), while Clairic returns values unknown in advance; the tutorial shows how to combine them.
- An interactive playground is available on DAIR.AI Academy for hands-on experimentation.
Related event: Cloudflare Releases Clairic Extraction Model; DAIR.AI Publishes Primer(2 posts)→
More from coding & agent
- A one-page project template with 9 fields for clearer scope and ownership — dSebastien · 2026-10-27
- Marketstack MCP Server brings stock market data APIs to AI assistants — modelcontextprotocol · 2026-10-27
- DMARKOFF MCP connector adds DMARC, SPF & DKIM analytics to AI assistants — modelcontextprotocol · 2026-10-27
- Kimi K3.1 Pro seen as an agent tradeoff play, not a benchmark winner — usamanoman · 2026-10-27
- Open-source plugin hard-blocks Qwen's hallucinated signed OSS URLs before agent tool calls — builderjaydub · 2026-10-27
- Fanout: open-source tool tames parallel Claude Code and Codex sessions before merging — crumb_ona_breadstick · 2026-10-27