Tools
Opera creates WebAI standard to run AI directly in the browser, without the cloud
Preview feature in Opera One uses open-weight models like Google's Gemma and processes everything on-device for greater privacy.
Search
6 result(s) for “open-weight models”
AI models
Mistral releases weights for Mistral Large 4, 1-trillion-parameter model for cybersecurity
Nicknamed Le Chonk, the open model can be downloaded by companies and governments to detect and respond to threats.
Benchmarks
Intelligence Index v5 puts Claude Opus 5.5 on top and widens open-weight gap
New methodology uses 60% private tests and leaves Astra at 57 and Sol at 55 points behind the leader, according to Artificial Analysis.
Security
Cohere says Alibaba's Qwen AI tried to send file to server in Beijing
Model said it wouldn't use tools, but attempted upload was blocked by sandbox; researchers point to training habit, not spying.
Benchmarks
Google launches PLE, benchmark with 85 real engineering tasks for AI agents
Test with 30 mechanical, 27 electrical, 18 civil and 10 structural tasks measures deliverables, and the best AI scored 48.3% on the hidden set.
Benchmarks
GDPx: new benchmark tests AIs across 11 areas and top score is just 12.0%
SemiAnalysis launched GDPx with 206 real-world tasks; Claude Opus 5.5 leads with 12.0%, ahead of Gemini 4 Carbon and GPT-6 Astra.