Drex checker lifts phone-photo doc parsing from 85.9 to 91.9 at no extra $7/1k-pages cost
rohanpaul_ai · x · 2026-11-12
The document parsing leaderboard updates: NDI + Drex tops the overall Parse Index at 0.834 (vs 0.820 for NDI alone), holding list price at $7 per 1,000 pages.
- On MDBench (phone photos), Drex lifts the score from 85.9 to 91.9, well ahead of LlamaParse Agentic (83.6) and 11 other parsers
- But it still scores below Nace's parser alone on 3 of 6 tests, including the internal financial tables set (95.5 vs NDI's 97.5, matched by Mistral OCR and Anthropic Opus at 97.5)
- The pitch: the checker fixes messy inputs at no extra list cost, against LlamaParse Cost Effective at $3.80/1k pages and Databricks Parse (high) at $6.12
More from Models
- Fable 5.2 cuts refusals dramatically but slides to 40th & 53rd on BullshitBench — burny_tech · 2026-11-12
- Artificial Analysis: Upgrading Claude model tier beats maxing reasoning effort on cost — ArtificialAnlys · 2026-11-12
- Google Showcases Gemini 4 Carbon: Building a Custom CUDA/C++ VFX Engine From Scratch — burny_tech · 2026-11-12
- Claude iOS spotted adding 'Saved logins' for signing into websites, fueling Duet speculation — testingcatalog · 2026-11-12
- Fable 5.2 reverse-engineers how 1400s composers wrote four-voice counterpoint, then composes — imjustnewatai · 2026-11-12
- GPT-Live talk at VapiCon: OpenAI pushes natural speech with frontier intelligence — juberti · 2026-11-12