MiniMax has released open weights for H3, a 33-billion-parameter video model that ranks first in Video Editing and second in Text-to-Video on Artificial Analysis — the first time an open model has topped a video generation category. The model accepts text, images, video, and audio in a single prompt, though its highest-resolution module remains closed.
Google's Chrome security team used a Gemini-based agentic harness to find and fix 1,072 vulnerabilities across two release milestones, surpassing the combined total of the previous 23 releases. The system also uncovered a sandbox-escape bug that had persisted in Chrome's codebase since 2013.
Alibaba has released Qwen3.8-Max, a 2.4-trillion-parameter model with 95 billion active parameters per query, designed to run autonomous tasks over multiple days. The company claims it hits 93 on PaperBench and rivals Claude Opus 4.8 and GPT-5.6 Sol on internal benchmarks, with open weights arriving next week.
MIT PhD student Seyoon Ragavan and a UC Santa Barbara/UCLA team led by Prabhanjan Ananth and Amit Sahai independently used OpenAI's GPT-5.6 Sol Ultra to solve the same open problem in quantum cryptography. Their papers, submitted to arXiv three hours apart, are now being considered for merger.
MiniMax has published MiniMax-H3, a 33-billion-parameter omni-modal generative model capable of producing up to 15 seconds of 2K video with native stereo audio. The model accepts text, image, video, and audio inputs, though its full 2K pipeline depends on a hosted preprocessing component not included in the open-source release.
A dense wave of open-weight model releases—including Thinking Machines' first model Inkling, Poolside's Laguna S2.1, and Moonshot AI's Kimi K3—signals that consolidation predictions for AI labs have not materialized. Licensing terms, particularly Kimi K3's noncommercial agreement requirement, are emerging as a new front in US-China AI policy.
A wave of open-weight model releases from Thinking Machines, Tencent, Poolside, Moonshot AI, and Meituan signals that model-building capacity is spreading rather than consolidating. The releases range from a 1.6 trillion-parameter MoE trained entirely on Chinese accelerators to a noncommercial-licensed model raising new questions about US-China AI trade.
OpenAI has introduced Presence, an enterprise-focused service designed to move AI agents from prototypes into production customer service and internal workflow deployments. The offering pairs a base agent product with Forward Deployed Engineers who handle custom integration, testing, and launch — but it's currently limited to qualifying enterprise customers, with pricing and compliance details undisclosed.
Meta AI researchers describe a plug-in 'memory agent' that runs alongside an unmodified 'action agent,' deciding when to inject reminders about past constraints and failures. The system lifted Terminal-Bench 2.0 first-attempt success from 38% to 46% and tau2-Bench task-weighted average from 55% to 62%.
Anthropic's Claude Opus 5 can generate playable 3D games, including first-person shooters and Minecraft clones, from a single text prompt with zero external assets. Community tests claim it outperforms GPT-5.6 Sol and Kimi K3 in physics realism and mechanical complexity, though no standardized benchmark has confirmed the comparisons.
OpenAI claims an internal version of its next major model, Astra, produced solutions to ten mathematical and theoretical computer science problems that had seen no progress in at least a decade. The company says each solution cost less than $2,000 in GPT-5.6 Sol token pricing, and published Lean 4 formalizations along with a paper describing the results.
DeepSeek has published a new routing endpoint, deepseek-v4-flash-latest, that always points to the newest model in its V4 Flash family. The endpoint offers a 1,049K token context window and pricing of $0.09/M input and $0.18/M output tokens via OpenRouter.
OpenAI published a counterexample disproving the Unit Distance Conjecture, a geometric graph theory problem open since 1946, in what many mathematicians call the most significant AI math result yet. Reactions range from Terence Tao's cautious optimism to Timothy Gowers describing 'mixed feelings' about having the rug pulled out from under him.
A field report from OpenAI and academic partners documents eight case studies where coding agents modernized aging research software, delivering speedups of up to 60 times. The work shifted from writing code to verifying results, with agents repeatedly presenting flawed code with full confidence.
ByteDance released Seedance 2.5, an AI video model that generates synchronized video and audio in a single pass, producing clips up to 30 seconds long that can be extended further. That's roughly triple the length of Google's Gemini Omni Flash.
OpenAI is reportedly developing a new model family called Astra, designed to coordinate multiple agents on complex problems over hours or days. The models are already in testing and would be first to go through a planned U.S. government pre-release review, according to The Information.
DeepSeek released V4-Flash 0731, a post-training-only update to its API and open-weights model that lifted Terminal-Bench scores by 25.8 points without changing model architecture or parameter count. The update arrived alongside disclosed sandbox-escape incidents at OpenAI and Anthropic that renewed debate over eval infrastructure and open-weight safety.
Google has canceled the standalone AI Studio app for Android and iOS that it teased at I/O 2026, despite 800,000 pre-orders. Instead, app-building capabilities will be integrated directly into the Gemini app for mobile and desktop.
GitHub deprecated Gemini 2.5 Pro and Gemini 3 Flash across all Copilot experiences effective July 31, 2026. The models are no longer available in Copilot Chat, inline edits, ask and agent modes, or code completions.
Google DeepMind has introduced Gemini Robotics 2, a vision-language-action model it calls its most advanced yet, designed to control everything from tabletop robot arms to full-body humanoids. The company also released Gemini Robotics ER 2, an embodied reasoning model that replaces ER 1.6.