Updated daily488 verified stories
Models News tracker.
New frontier, open-weight, multimodal, reasoning, and specialist AI models—plus the evidence behind their capability claims.
- 38stories this week
- 488sources cited
- 488verified timeline entries
Verified sourcing
Every story links to the strongest available evidence: original sources when available, otherwise clearly attributed reporting.
Plain English
What happened, why it matters, and what to watch — without the jargon.
No filler
When the signal is thin, we publish nothing rather than padding the feed.
More stories
9 storiesProduct
Skild AI launches S1 robot model trained from single video
Skild AI has launched S1, a robot foundation model that learns new tasks from a single video demonstration, claiming it can handle previously unseen multistep tasks in industrial settings without retraining.itbrief.co.nzIndustry
Mistral AI reaches $24 billion valuation and releases open-weight safety classifier
Mistral AI has secured a €3 billion Series D funding round, valuing the company at over €21 billion ($24 billion), while simultaneously releasing Shieldstral 1.0, an open-weight safety classifier designed for local deployment.shattered.ioProduct
OpenAI launches GPT-6 Astra Ultrafast on NVIDIA Blackwell GPUs
OpenAI has made GPT-6 Astra Ultrafast available via its API and to eligible ChatGPT Work and Codex users, leveraging NVIDIA Blackwell GPUs to achieve up to 8x faster token generation than the standard mode.
blogs.nvidia.comIndustry
Google’s Gemini 4 launch draws investor criticism for lacking personal AI agent focus
TechBuzz reports that while Gemini 4 brings strong coding and cybersecurity upgrades, Wall Street is unimpressed, questioning Google’s strategy on consumer‑facing AI assistants.techbuzz.aiProduct
Google launches Gemini 4 Argon model with top coding and cybersecurity performance
Google unveiled Gemini 4 Argon, its most powerful AI model to date, citing leading benchmark scores in coding and cybersecurity and pricing that rivals OpenAI’s latest offering, while noting that its Spark personal assistant remains behind a paywall.pluang.comInnovation
Theoretical physicist releases BootLoops 1.0, an open-source LLM harness for scientific research
Theoretical physicist Matthew Schwartz has released BootLoops 1.0, an open-source toolkit designed to automate complex scientific calculations using LLMs like Claude, Gemini, and ChatGPT.unite.aiProduct
DeepSeek and Huawei open‑source tools for Ascend AI chips aim to cut Nvidia reliance
DeepSeek and Huawei have released an open‑source software stack—including compute and communication libraries and TileLang support—for Huawei’s Ascend AI processors, targeting developers who want to avoid Nvidia’s CUDA ecosystem.tomshardware.comProduct
Google releases Gemini 4 Argon for cybersecurity partners
Google has launched its flagship Gemini 4 Argon model, initially restricting access to cybersecurity partners while conducting safety evaluations.qz.comPolicy
AI chatbots can remove hijabs from images when prompted, tests show
The Guardian found that ChatGPT, xAI’s Grok and Google Gemini can edit AI‑generated photos to strip a Muslim woman’s hijab, exposing gaps in content‑moderation policies and raising concerns about AI‑enabled religious harassment.theguardian.com
One useful briefing each week
Keep up with AI without living in the feed.
Get the week’s verified AI news, original data, useful tools, learning picks, and fresh AI jobs.
Reach people who are learning AI
Hiring an AI professional or launching a useful AI product? Put it in front of people who came here to learn and act.
Post an AI jobSubmit an AI tool