NVIDIA Launches Nemotron 3.5 Super VL and Bolt
According to NVIDIA, Super VL is a 120-billion-parameter open model for agents, and Bolt 74B runs locally on NVIDIA hardware. Both are available now; Bolt E4B ships October 23.
Artificial intelligence, explained in plain language.
According to NVIDIA, Super VL is a 120-billion-parameter open model for agents, and Bolt 74B runs locally on NVIDIA hardware. Both are available now; Bolt E4B ships October 23.
According to Microsoft AI, MAI-Base-2 beats every open-weight base model it tested. Post-training is still underway, with weight licensing planned for trusted Foundry partners.
The five AI stories that mattered from October 12 to 14: Gemini 4 Argon, The X Subscription, Grok Bot 0.70.2, Muse Wonder and small-model self-replication, with prices and availability.
A first look at ChatGPT Images 2.5 from OpenAI’s Help Center: what Free includes, when you need Images with thinking, editing, templates and privacy.
SpaceXAI’s new voice model tops the Artificial Analysis speech-to-speech index, replies in under a second and keeps the same price as its predecessor.
A comparison from official docs: prices, automation, working in websites, teams, model choice and privacy, plus which fits business owners, writers and developers.
In the new update, every bot gets an inbox of its own, reads X without a connected account and can pay, and it checks for your approval before anything irreversible.
Instead of guessing from your likes, Facebook now asks you to tell it what you are into, and large language models turn that into real-time recommendations.
From 2,500 merges a day in May to over 30,000 in September: tools built for humans could not keep up with agents, so Anthropic built a new system in five months.
A first look at Gemini 4 Argon from official info: Pro and Ultra access, API pricing, the 1M unified token limit, Tasks in the Gemini app and Google’s benchmarks.