Enterprise AI gap widens as open-weight models mature into production-ready alternatives

A large-scale study testing 34,198 real-world skills reveals that AI agent performance drops drastically when moving from curated benchmarks to realistic conditions. Claude Opus 4.6 saw pass rates fall from 55.4% with hand-selected skills to 38.4% in truly realistic scenarios, while weaker models like Kimi K2.5 actually perform below their no-skill baseline.

Arcee AI has released Trinity-Large-Thinking, a 400-billion-parameter open-weight reasoning model with a mixture-of-experts architecture that activates only 13 billion parameters per token. The model matches Claude Opus 4.6 on agent benchmarks like Tau2 and PinchBench but lags on general reasoning tasks. The company spent approximately $20 million—roughly half its total venture capital—to train the model on 2,048 Nvidia B300 GPUs over 33 days.

MiniMax has released MiniMax-M2.7, a 229-billion parameter model that uniquely participates in its own evolution during development. The model achieves 66.6% medal rate on MLE Bench Lite and 56.22% on SWE-Pro benchmarks, with native support for multi-agent collaboration and complex tool orchestration.

Google released Gemma 4, an open-source multimodal model that runs entirely on smartphones without sending data to the cloud. The E2B and E4B variants require just 6GB and 8GB of RAM respectively and can autonomously use tools like Wikipedia, maps, and QR code generators through built-in agent skills. The model is available free via the Google AI Edge Gallery app for Android and iOS.

Researchers introduced ProactiveBench, a benchmark testing whether multimodal language models ask for help when visual information is missing. Out of 22 models tested—including GPT-4.1, GPT-5.2, and o4-mini—almost none proactively request clarification, instead hallucinating or refusing to respond. A reinforcement learning approach showed models can be trained to ask for help, improving performance from 17.5% to 37-38%, though significant gaps remain.

Anthropic has launched Ultraplan, a new feature for Claude Code that offloads programming task planning to the cloud. The feature enables developers to initiate planning jobs from the terminal while the planning executes in the browser, supporting inline comments, emoji reactions, and revision requests on individual plan sections.

Liquid AI has released LFM2.5-VL-450M, a refreshed 450M-parameter vision-language model built on an updated LFM2.5-350M backbone. The model features a 32,768-token context window, supports 9 languages, handles native 512×512 pixel images, and adds bounding box prediction and function calling capabilities. Performance improvements span both vision and language benchmarks compared to its predecessor.

Vice President JD Vance and Treasury Secretary Scott Bessent held a call with leading tech CEOs including Anthropic's Dario Amodei, OpenAI's Sam Altman, and Google's Sundar Pichai to discuss AI model security and cyber attack response. The meeting occurred one week before Anthropic released its Mythos model, which has major cybersecurity implications and raised concerns at the Federal Reserve and among top U.S. banks.

Google has redesigned AI Mode's prompt interface with a bottom sheet layout on mobile and expanded its agentic restaurant booking feature to 8 new markets including the UK, Canada, and Australia. The update rolls out to stable channel on Android and iOS.

Tencent has released HY-Embodied-0.5, a family of foundation models designed specifically for embodied AI and robotic control. The suite includes a 2B-parameter MoT (Mixture-of-Transformers) variant with only 2.2B activated parameters during inference, and a 32B model that claims frontier-level performance comparable to Gemini 3.0 Pro, trained on over 200 billion tokens of embodied-specific data.

Federal Reserve Chairman Jerome Powell and Treasury Secretary Scott Bessent met with major U.S. bank CEOs this week to assess cyber risks from Anthropic's newly released Mythos model. Anthropic deployed Mythos in limited capacity due to concerns that hackers could exploit its capabilities.

Microsoft has begun removing Copilot branding and buttons from Windows 11 applications including Notepad and Snipping Tool, replacing them with generic icons like a pen symbol. The underlying AI-powered features remain functional but are now labeled as "writing tools" rather than Copilot. This follows user complaints about forced Copilot integration and inconsistent experiences across apps.