Sparrow 0.6.0 open-source doc extraction reads skewed scans into rows keyed by the page's own column headers
andrejusb · x · 2026-11-11
Katana ML's open-source document extraction tool Sparrow 0.6.0 introduces table mode: no query or JSON schema needed — point it at a skewed, scanned lab report and it returns rows keyed by the page's own column headers (Test, Result, Flag, Unit, Reference, Prev), leaving empty cells empty. Runs on your own infra or a private cloud GPU.
The hosted version offers:
- PDF/image upload (≤5MB, deleted right after inference), schema-driven extraction with typed JSON fields and validation
- Table-only mode for dense tables, financial reports, lab results, and statements
- Standard/Advanced Vision LLM tiers; without a key, 30 calls per 6 hours on 3-page documents
Broader capabilities cover multi-page PDFs, page classification, bounding-box annotation, and a Sparrow Agent chaining Vision LLM extraction with Text LLM reasoning and field validation.
Related event: Sparrow 0.6.0 Adds Table Mode for Schema-Free Extraction From Skewed Scans(2 posts)→
More from coding & agent
- n8n Labs launches open-source agent Nod: zero false memories in 216 injection attacks — Aiden_Tech_Ai · 2026-11-11
- Anthropic launches self-hosted Claude Managed Agents in private beta — Claude (Anthropic) · 2026-11-11
- Anthropic launches self-hosted Claude Managed Agents in private beta — danshipper · 2026-11-11
- Codex Remote Chats Now Integrated into All ChatGPT iOS Chats — SIGKITTEN · 2026-11-11
- Cloudflare proposes identity, reputation, accountability frameworks for AI agents on the web — bibryam · 2026-11-11
- banteg launches Sokoban Eval: trilogy solved in 3,000-3,064 keys, last 64-key gap unproven — banteg · 2026-11-11