Mistral Pixtral Large 2.5 reaches SOTA open multimodal reasoning on MMMU-Pro and MathVista-Hard

[AI-NEWS]

Date: 2026-02-25

Content : Pixtral Large 2.5 (open-weight) sets new open-model records on MMMU-Pro (visual reasoning) and MathVista-Hard (math + vision), surpassing previous open SOTA by 9–14 points while maintaining efficient inference on 8×A100 or consumer multi-GPU setups.

Keywords : multimodal reasoning, MMMU-Pro, MathVista, Pixtral Large 2.5, open-weight vision-language model

20 2026-02-25

A Complete Guide to Geoffrey Hinton's Methodology: The "Father of AI Deep Learning"

OpenAI o4-proto-4 achieves 82% autonomous resolution on full SWE-Bench Verified + Multi-File

New

OpenAI o4-proto-4 achieves 82% autonomous resolution on full SWE-Bench Verified + Multi-File Google DeepMind AlphaFold 4 predicts all known protein complexes with near-experimental accuracy xAI Grok-5.1 integrates real-time multi-agent debate for improved truthfulness and calibration Anthropic Claude 5 Sonnet demonstrates 14-day autonomous scientific literature review + hypothesis generation Mistral Pixtral Large 2.5 reaches SOTA open multimodal reasoning on MMMU-Pro and MathVista-Hard A Complete Guide to Geoffrey Hinton's Methodology: The "Father of AI Deep Learning" After Reading All Blogs by Peter Steinberger, Founder of OpenCLaw: Extracting the Core Methodology The three stations in Wuhan have opened the Light Travel service Wuhan Spring Festival Lantern Show is coming soon! How to apply for China's barcode invitation letter?