Mistral Large 4 on STEMBench: Strong in Engineering
Mistral Large 4 reached 2.501 on STEMBench, showing high marks in engineering and astronomy while placing sixth overall among ten frontier AI models.
Claude Mythos 5 Designs Leaner Quantum Circuits
Anthropic says Claude Mythos 5 beat human benchmarks on quantum circuits with Harvard and QuEra. The gains are theoretical, with no physical hardware tests.
OpenAI Opens Rosalind Academic Plan to US Labs
OpenAI launched the Rosalind Academic Plan, granting US labs 30 free ChatGPT Business seats and GPT-Rosalind access for a year, with no API access.
Hermes Agent Adds NVIDIA Nemotron 3.5 Support
Nous Research integrated Hermes Agent with NVIDIA Nemotron 3.5 Super VL, reaching 63.44 on Hermes Bench and up to 5x higher throughput.
Mistral AI and CERN Build Large Physics Model
CERN and Mistral AI partner to build the Large Physics Model, using simulated LHC collisions to handle surging particle physics data.
Meta Omnilingual ASR Named to TIME's Best Inventions 2026
Meta Omnilingual ASR made TIME's Best Inventions of 2026. The open speech model covers 1,600 languages and aids community health clinics in Nigeria.
Cursor Adds Meta Muse Wonder with 46.1% Record
Cursor integrates Meta's Muse Wonder reasoning model, hitting 46.1% on CursorBench with a 1M token window and $20 per million output tokens.
Perplexity Open-Sources SPACE Sandbox
Perplexity open-sourced SPACE, its Firecracker microVM sandbox for AI agents, separating system root access from API credentials and network gateways.
OpenAI Releases GPT-Live-1.1 with New Accents
OpenAI rolled out GPT-Live-1.1 to ChatGPT Voice with new regional accents, natural search pauses, and the compact GPT-Live-1.1 Quip model.
Nvidia Nemotron 3.5 Bolt Fits 74B Model Into 24GB
Nvidia launched Nemotron 3.5 Bolt, a 74B MoE model compressed via 2-bit quantization to run agentic coding and browser workflows on 24GB hardware.