Skip to content
trckr.ai

Loading…

News

The last 24 hours

Anthropic priced its IPO at $83 a share, the largest on record, raising $81.3 billion. Google released Gemini 4 Neon, its fourth new Gemini model in about seven weeks, and OpenAI reported that agents now complete 46% of its researchers' multi-hour tasks on their own.

133 stories ranked · updated Nov 19, 15:04 UTC · refreshes during the day

  1. 01

    Anthropic prices record IPO at $83 a share, above its initial range

    Anthropic priced its IPO at $83 per share, above its $70 to $78 range, and raised $81.3 billion, the largest IPO on record. Critics question its safety branding, and an early investor says Frontier model: The most capable models of the moment, from the handful of labs at the cutting edge. labs should slow down.

  2. 02

    Google releases Gemini 4 Neon at $3 and $15 per million tokens

    Gemini 4 Neon is available in the API: How developers plug a model into their own apps and pay per use. A model that's "in the API" can be used by programmers, even if it isn't in a chat app yet., Antigravity and the Gemini app, with a A 1M-token context window: It can keep about 750,000 words, or 2,500 book pages in mind at once: the question, any files you share, and its own replies.. Artificial Analysis says it trails only Claude Fable 5.2 and GPT-6.2 Astra on its Intelligence Index.

  3. 03

    OpenAI says agents now finish 46% of researchers' 2-to-4-hour tasks alone

    OpenAI reports that Agent: An AI that doesn't just answer but takes steps on its own to finish a task: searching the web, running code or using apps. complete 46% of its researchers' 2-to-4-hour tasks on their own, up from 23% in March. OpenAI also says its models improved power, performance and area by up to 14% on AMD's Flagship: A lab's best and usually most expensive model, the one it leads with. GPU: The chip that runs AI models. Bigger models need more and larger GPUs, which is why the biggest ones live in data centres. IP.

  4. 04

    Microsoft unveils seven in-house MAI models and replaces some outside models

    Microsoft's AI division MAI introduced seven models for code, security, images, language and routing at Ignite. MAI Code 2 reaches 98% of GPT-6.1 Sol's performance at 40% of the cost, according to Microsoft.

  5. 05

    Jensen Huang steps down at Nvidia; Colette Kress named interim CEO

    Nvidia founder Jensen Huang stepped down as CEO last week amid several scandals. Interim CEO Colette Kress told staff at the first all-hands since then to expect pay raises and more equity in December.

  6. 06

    US remote-access chip rules hit Chinese AI labs; Moonshot cancels $150M contract

    Tencent President Martin Lau says new US remote-access rules on Chinese chip access will delay new AI model training. Moonshot AI reportedly canceled a $150M contract to rent Nvidia chips in an Indonesian data center after the US rules.

  7. 07

    Google's Genie 4 world model rolls out to AI Ultra subscribers

    Google DeepMind's Genie 4 generates interactive 3D worlds that can be changed in real time through text input. It is available to Google AI Ultra subscribers through Project Genie, which also adds a community gallery.

  8. 08

    Arizona opens criminal investigation into OpenAI and other AI companies

    Arizona's attorney general has opened a criminal investigation into OpenAI and other AI companies over what it calls the apparent infiltration of banks by rogue AI Agent: An AI that doesn't just answer but takes steps on its own to finish a task: searching the web, running code or using apps., according to Reuters.

    OpenAITechmeme ↗
  9. 09

    Newsom proposes broad California powers to regulate AI industry

    California Governor Gavin Newsom unveiled a sweeping proposal giving the state broad powers to regulate the AI industry. It stops short of requiring developers to build kill switches.

Releases

New models today

NVIDIASpecialist

MinorOpen

C-RADIOv4-H-AV

NVIDIA released a free-to-download search tool that finds driving scenes in large video collections by text or example picture, for self-driving car developers.

TencentVideo

MinorOpen

WeVisForensics

Tencent released a free research tool that helps detect AI-made videos by checking how they look, move, and sound; it is built on Alibaba's Qwen model and meant for researchers.

NVIDIAText & chat

MinorOpen

Qwen3.8-27B

NVIDIA's slimmed-down copy of Alibaba's Qwen3.8 27B chat model, made to run faster on NVIDIA's newest graphics chips. Free to download for developers building chatbots and AI helpers.

Get the news every evening.

Turn on the daily digest for any channel: email, Telegram, Discord, push or ntfy.

Set up the digest

Archive

Earlier editions

Free alerts

Know the moment a new model drops.

Pick one or several. Each has its own settings for what you hear about.

trckr is free and has no ads. If it's useful to you, you can chip in.