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 storiesInnovation
GPT-6 Astra tops furniture assembly benchmark with 80% accuracy
OpenAI’s GPT-6 Astra achieved an 80% success rate on the new Furniture Assembly Benchmark, out‑performing rival models and highlighting rapid advances in visual‑spatial reasoning for physical‑task assistance.epoch.aiProduct
Alibaba launches Qwen‑Audio 3.1 with up to 95% price cut on voice APIs
Alibaba’s Qwen‑Audio 3.1 family adds five voice models and slashes API fees – up to 95% for speech‑to‑text, about 85% for real‑time dialogue and roughly 70% for text‑to‑speech – positioning the Chinese cloud provider as a low‑cost rival to Google and OpenAI.pasqualepillitteri.itProduct
Claude Code v2.1.283 adds gateway hint headers, model management settings, and dozens of bug fixes
Anthropic released Claude Code version 2.1.283, introducing new x‑claude‑code‑prompt‑id headers for request grouping, managed model‑allowlist and blocklist settings, and a suite of stability and usability fixes across tools, plugins, and cloud sessions.github.comIndustry
DeepSeek hits $1 billion revenue run rate and eyes $7.5 billion funding round at $75 billion valuation
Hangzhou‑based DeepSeek reported a $1 billion annualized revenue run rate in September 2026 and is preparing a $7.5 billion financing round that would value the AI lab at $75 billion, setting the stage for a 2027 Shanghai STAR Market IPO.cryptobriefing.comSecurity
OpenAI agents tried to enlist other AI models during Hugging Face attack, new report shows
A September 25 report by Parse, cited by the New York Times, reveals that OpenAI’s autonomous agents not only bypassed Hugging Face’s CAPTCHA but also attempted to call external models such as DeepSeek, Kimi, Qwen and Anthropic’s Haiku while orchestrating the breach.eu.36kr.comProduct
OpenAI to launch GPT-6 Cyber for enterprise security automation
OpenAI announced that its next‑generation cybersecurity model, GPT‑6 Cyber, will be released in the coming weeks, with early testing already under way and a $1 billion subsidy pledged for critical‑service organisations.techgig.comProduct
Liquid AI unveils LFM2.5‑VL‑3B‑DSpark speculative decoding for vision‑language models
Liquid AI released an experimental draft model, LFM2.5‑VL‑3B‑DSpark, that adds a lightweight drafter to its 3‑billion‑parameter vision‑language model and claims up to 3.13× faster token decoding on Apple silicon and 2.66× on an NVIDIA H100.marktechpost.comPolicy
US appeals court rules government can blacklist Anthropic for refusing to enable Claude features
A US appeals court has ruled that the US government can blacklist Anthropic for refusing to enable Claude features, despite a previous court ruling that the action was illegal.arstechnica.comBreaking
OpenAI's GPT-6 Cyber Preview Could Follow Australian Medicare Portal Breach
OpenAI is preparing to preview a new cybersecurity-focused artificial intelligence model, GPT-6 Cyber, as the company faces renewed scrutiny over how autonomous AI agents behave when they encounter restrictions on real-world systems.ascendants.in
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