OpenAI releases GPT-6.2 Astra and Sol with a 20% price cut on Astra, Nvidia opens a chemistry dataset and post-training recipes for science agents, Isomorphic Labs raises $5 billion, and Nvidia reports $115.8 billion in quarterly revenue under its interim CEO.
GPT-6.2 Astra costs 20% less than GPT-6 Astra, at $8 per million input tokens and $40 per million output tokens. Sol keeps its $2 and $10 pricing and comes close to Astra on coding at a fraction of the cost.
Nvidia's Nemotron for Science combines an open vision-language model, about 50,000 chemistry questions in LitmusBench v0.2 and post-training recipes for scientific agents.
Isomorphic Labs raised $5 billion in a Series C co-led by Greenoaks and Sequoia, to move its first AI-designed drugs toward clinical trials.
Nvidia's revenue doubled to $115.8 billion, and its board said it will not bring back Jensen Huang as chief executive.
Google's fast Gemini 4 Hydrogen model, ByteDance's cheaper Seedance 2.5 Flash video model, Nvidia's CEO transition and SpaceX's $74 billion data-centre plan, with tools to test.
Special Edition · Chatbots · Code Generation · Finance AI
A Special Edition round-up: new Gemini, Claude and Qwen tiers, AI listings from Nasdaq to Hong Kong, UK backing for Fractile, ChatGPT Wallet and the Microsoft 365 outage.
Claude adds sheets, whiteboards and data stories, Tencent opens Yuanbao Search, Anthropic starts its IPO roadshow and AI tokens hit Singles' Day shelves. What each means for your tools.
MiniMax releases open-weight H3 Image for 4K generation, OpenAI lets companies claim profiles in ChatGPT, and China reports 3.1 million PFLOPS of compute as Alibaba plans a T-Head listing.
Anthropic ships Claude Fable 5.2 and Mythos 5.2, Meta launches its low-cost Muse Ember model, and AI labs look to gas generators and satellites for computing capacity.
Google's Gemini Omni 1.2 Flash makes 15-second AI video at $0.07 a second, OpenAI closes $11 billion at a $1.25 trillion valuation, and SMEE shows new lithography machines.
Special Edition · Chatbots · Data Analysis · Finance AI
A Special Edition round-up: GPT-6 Cyber, Tencent Hy4 and Claude Personal Finance, the Marvell–Lumentum deal and AMD's data-centre surge, METR's 1.9x finding, and new AI policy in China and the UK.
OpenAI's Images 2.6 adds animated GIFs and SVG export, Moonshot launches Kimi Hosted Agents, China's MIIT tightens rules for phones with built-in AI agents, and chip stocks slide after the US midterms.
Moonshot releases the weights of Kimi K3.1 Pro, Tencent makes Hy4 generally available, Chinese tech firms sign an agent self-regulation pact, and OpenAI buys Git tooling maker Pierre for Codex.
Apple Intelligence enters China beta with Qwen, DeepSeek Creator Mode, AI industry research and video workflows, with practical tools and category guides.
Kimi launches a game development solution that works inside Unity, Unreal, Blender and Godot, Tencent shows Hyra 2.0 for AI-run research, and Anthropic's amended S-1 reports $17 billion in quarterly revenue.
Meta builds a humanoid robot prototype with China's Fourier, DeepSeek shows a Router-Worker design for always-on agents, major AI companies plan to honour a common robots.txt opt-out for training, and Kling AI closes a ¥19.8 billion raise.
Special Edition · Chatbots · Code Generation · Finance AI
A Special Edition round-up: GPT-6.1 Terra halves prices, Mistral opens Large 4, Anthropic files for a Nasdaq IPO, Meta and SpaceX line up record financing, Google runs Gemma 4 in orbit, and new energy and publishing rules in China and the UK.
AgiBot previews GO-3, a robot model that keeps learning after deployment, Tencent links its office apps under one AI system called Buddy One, Guoxing Aerospace clears its Hong Kong listing hearing, and Marvell enters talks to buy Lumentum.
Google launches Gemini 4 Carbon with access first for vetted security teams, Apple's iPrefBench shows phone agents act well but personalise poorly, and Alphabet's quarter pairs record cloud growth with $51 billion of capital spending.
Google DeepMind's Genie 4 generates interactive worlds with sound in real time, Nvidia invests $3 billion in Meta as Muse passes 10 million users, and China publishes its first national standard for data processing units.
Tencent open-sources Hy-SimulMT for real-time Chinese–English interpretation, Anthropic's S-1 shows $16.3 billion of first-half revenue and heavy reliance on cloud partners, and SpaceX and Meta report the results and funding behind their AI build-outs.
CAICT reports Chinese AI token use reached 500 trillion a day while average prices nearly doubled, China's State Council issues guidelines to soften AI's impact on jobs, and PICC launches the country's first AI risk insurance products.
Special Edition · Chatbots · AI Security · Code Generation
A Special Edition round-up: Claude Duet and OpenAI on Google Cloud, China's new national data group, Nvidia chip orders and DeepSeek's funding, the AI-found PHP2Shell flaws, OpenAI's offline training, and new AI policy in China and the UK.
Recursive open-sources Flint, a small-active-parameter coding model built largely by its own AI research system, China's MIIT and seven other departments set out AI's role in strengthening industrial foundations, plus PHP2Shell and Claude Duet in brief.
Nvidia's DSX POD ships AI data centres in containers, AMD and Intel merge their CPU optimisations for PyTorch into oneDNN, China sets out how to fuse science with industry, and OpenAI buys drug-development start-up Orchestra Bio.
· Code Generation · Speech to Text & TTS · AI Security
DeepSeek releases DeepInfer, the framework behind its production service, Ant Group open-sources five Ming-3.0 speech models, and the US Department of Homeland Security orders civilian agencies to remove DeepSeek products.
DeepSeek updates V4.1 Flash and opens paid Search and Fetch APIs, Tencent open-sources Hy Memory 2.0 to give agents long-term memory, Sugon shows its fully domestic Sugon 8000 supercluster, and Anthropic lines up $63 billion of chip financing.
Nvidia doubles the speed of its DLSS 5 neural rendering model, ModelBest releases MiniCPM-V 4.7 to run a 35B MoE model on phones, IDC finds Nvidia's China AI chip share fell to 34%, and Character.AI files for bankruptcy with Disney as the opening bidder.
Meta launches its first frontier models, Muse Wonder and Wonder Full, Alibaba's Qwen team shows QwenGyre for training long-running agents, OpenClaw adds skill search, MCP apps and full backups, and Microsoft unveils the Maia 300 inference chip.
Microsoft launches Decision-1, a fast scoring model built on Qwen, TypeSafe raises $870 million for its non-text model Jev, Anthropic offers free AI vulnerability scans to open-source maintainers, and China sets out how to develop new quality productive forces.
Get HAI Daily in your inbox
The day's AI news, what it changes, and which tools to try.