Datalab fast-mode OCR beats Opus 5.5 and GPT-6.1 Sol at 1/4 to 1/68 the price
VikParuchuri · x · 2026-11-12
Datalab founder Vik Paruchuri announces its fast-mode OCR now scores 91.6 on OmniParseBench (16,288 tests, 2,937 pages), beating Claude Opus 5.5 (90.0), GPT-6.1 Sol (91.5), and Mistral OCR (89.5).
- Cost: $1 per 1,000 pages, 1/4 to 1/68 of competitors' prices (Opus 5.5: $68.45)
- Speed: median 3.6s per page, 1/3 faster than balanced mode
- Only Gemini 4 Argon/Carbon (95.8/95.6) and Gemini 3.8 Flash score higher — at 37x+ the cost and 6x+ slower
Benchmark is open-sourced at github.com/datalab-to/omniparsebench (Apache-2.0).
Related event: Datalab's Fast Mode OCR Nears Gemini Accuracy at a Fraction of the Cost(2 posts)→
More from Models
- 11 LLMs battle in 1,168 Cambridge Union debates; Claude Fable 5.2 dominates at 209-14 — AymericRoucher · 2026-11-12
- Claude app widget bugs: edits invisible to Claude, mobile remote control broken — petergyang · 2026-11-12
- Allen-Zhu's new paper: Canon layers boost reasoning 2x, even random convolution weights help — teortaxesTex · 2026-11-12
- Ricursive releases open-weight Ripple-65B, claiming US open-weight SOTA on TerminalBench 2.1 — StanfordAILab · 2026-11-12
- A criminally underrated resource comparing tool-calling reliability across models and providers — ricklamers · 2026-11-12
- Claude's world knowledge impresses: model guesses user identity from one prompt — adonis_singh · 2026-11-12