Datalab ships Chandra 2.6: beats frontier models on OmniParseBench at 94.1, 2.5x faster
VikParuchuri · x · 2026-10-28
Datalab released Chandra 2.6 across all API modes at unchanged pricing. On OmniParseBench, its accurate mode scores 94.1, ahead of Gemini 3.8 Flash (92.7), GPT-6.1 Sol (91.4) and Claude Opus 5.5 (90.7). Median per-page latency on a fixed 1,187-page set dropped from 17.0s (Chandra 2.4) to 6.7s — 2.5x faster and 37% faster than 2.5 — while failed requests fell from 1 in 6,400 to 1 in 112,000. No config changes needed for existing users.
Related event: Datalab's Chandra OCR 2.6 beats frontier models at 5-15x lower price(3 posts)→
More from Models
- Open model Flint hits #2 trending on Hugging Face, team outlines next steps — josh_tobin_ · 2026-10-28
- YacineMTB rails: even the best LLMs are 'complete dogshit' and he's tired of it — yacineMTB · 2026-10-28
- Omni-InteractScope benchmarks proactive interaction in continuous omnimodal streams — Qwen · 2026-10-28
- Mistral releases Timestral time-series foundation models, tops GIFT-Eval — mistralai · 2026-10-28
- Gemini 4 Argon Hits 1.73% WER on OCR Benchmark, Surpassing Human Experts — gabriberton · 2026-10-28
- Meta's Omnilingual MT covers 1,600 languages, 1B–8B models rival 70B LLMs — kalyan_kpl · 2026-10-28