CompaniesClaude Code now works while your laptop is closed.
· 3 min read
ExplainedWhen the code assistant starts doing the work.
· 4 min readTrending
New AI models this week: Kimi K3.1 Pro and DeepSeek V4.1 Flash lead
Moonshot's Kimi K3.1 Pro leads a week of three new API models and open…
DeepSeek API: Models, prices in CNY, free chat app and how to get a key
The DeepSeek API serves two models, deepseek-flash and deepseek-v4-pro, in OpenAI and Anthropic formats,…
Advertisement
Latest News
Nvidia’s open 6G tools pass 1 million downloads, with 45.6% less RAN power in a test
Nvidia says developer adoption of Aerial and Sionna grew 10x in a year, and shares research results from King's College London, SoftBank, ETH Zurich,…
AMD’s PrimusClaw keeps agents running on your own GPUs: 40,000 sessions so far
AMD released PrimusClaw, an MIT-licensed runtime that hosts agent sessions, harness and sandboxes in your own Kubernetes cluster; AMD says it has…
ByteDance’s Cola2 trains on 346B tokens, 60.1 average against 75.9 for Qwen3
ByteDance Seed put Cola2 on Hugging Face, a 9B base model that plans in compressed latents before writing tokens. It averages 60.10…
Sparse autoencoders: How researchers split a model’s activations into readable features
A sparse autoencoder rewrites a language model's internal activations as a few active features out of millions. Here is how it works,…
SoftBank’s credit risk hits a four-year high as AI fears spread
Reuters reports that SoftBank's credit risk has jumped to a four-year high as AI fears unsettle markets, a day when CNBC reported…
JFrog lifts 2026 revenue view to $680M as AI code floods software supply chains
JFrog's Q3 revenue rose 31% to $179.2 million, and the company lifted its full-year 2026 revenue outlook to $680 million-$682 million, citing…
Featured
AI API pricing tracker: What every major model costs per million tokens
What every major AI model costs through its API, per million input and output tokens, updated…
Free AI video generators in 2026: What you get without paying and where the limits are
Few AI video generators are free in the full sense: most give a one-time credit pack,…
Kimi can now drive Blender and CAD tools to build 3D scenes and games
Moonshot AI says Kimi 3D Make links the assistant to Blender and CAD tools to make…
New AI models this week: Meta’s Muse Wonder and Google’s Gemini 4 Argon
Meta's Muse Wonder and Google's Gemini 4 Argon lead a week of five priced releases, from…
New AI models this week: Kimi K3.1 Pro and MiniMax M3.1 Flash lead
Moonshot's Kimi K3.1 Pro and MiniMax's cheap M3.1 Flash lead the week, alongside Recursive's Flint, Mistral's…
Cursor AI pricing in 2026: Plans from $0 to $200, models, limits and alternatives
Cursor has seven plans, from a free tier to a $200 Ultra plan, with one weekly…
OpenAI
GPT-Live can now place phone calls: OpenAI opens outbound SIP to all
- OpenAI’s GPT-6 Cyber goes to vetted defenders first, not to the public API
- Claude Code’s product lead Cat Wu is moving to OpenAI, Business Insider reports
- ChatGPT can now make GIFs, and OpenAI says it serves 4B images a week
Google’s Omni 1.2 Flash makes 15-second AI video 30% cheaper
- Gemini can now turn a photo of your sketch into a working app
- Google lets schools switch off AI, but only by losing Search
- Schools that switch off Google’s AI will lose Search too
Anthropic
METR: Anthropic’s models now improve about twice as fast, mostly thanks to AI
- Claude now takes 29% of UK AI subscription spending, up from 7% in December
- Claude helped solve a bacterial vault structure in 34 hours, Anthropic says
- Anthropic reportedly pays AP for news content, its first such deal
Advertisement
Top categories
AI Tools & Guides
What the tools do, what they cost, and how to start
ChatGPT prompts for photos: Editing, pictures of yourself and how image editing works
Upload a photo, say what to keep and what to change, and ChatGPT edits it in the same…
AI agents platform guide: What builders do and how to pick one in 2026
An AI agents platform lets you build, run and monitor software that pursues a goal with tools. Here…
Claude vs ChatGPT: Plans, prices and which one fits which job
Both assistants start free and cost $20 a month at the first paid tier. The differences lie in…
Qwen AI explained: Alibaba’s models, Qwen Chat, Qwen Code and the image editor
Qwen is Alibaba's AI model family: open-weight models you can download, paid cloud models, an image generator and…
Advertisement
More News
Explore more stories from Aiholics News
Airbnb’s own AI travel agent is due in 2027, and it won’t be a chatbot
Airbnb's Q3 shareholder letter sets the first half of 2027 for its AI travel agent, which swaps the chat window…
Ai2’s ManiFlow-tb robot arm model finishes about 57% of block tasks in simulation
Ai2's open ManiFlow-tb drives a simulated Franka arm and completed 54% to 59% of block tasks across four runs. Results…
Ai2’s block-stacking robot planner scores 97.5% on easy tasks, 14.5% on hard unseen ones
Ai2's Molmo2-4B-tb is a 4.5B-parameter model that looks at a camera image and decides the next block-stacking step. Its success…
Viggle Remake puts video recasting on a 12 GB GPU, with a license catch
Viggle has published Remake, a video-to-video model that recasts people and changes scenes in existing footage, with LoRA weights and…
AMD profiles DeepSeek-V4-Flash training: 556 ms to 239.8 ms per step
AMD's profiling walkthrough for Primus takes an eight-layer DeepSeek-V4-Flash proxy from 556 ms to 239.8 ms per step and shows…
Perplexity’s new ‘teammates’ work on their own and can message your coworkers in Slack
Perplexity says its new teammates take on jobs proactively, collaborate with each other and can message coworkers in Slack, but…
Nvidia’s VFlash decodes robot trajectories in one pass, 5.1x faster
Jim Fan says VFlash outputs whole boxes and trajectories at once, making robot inference 5.1x faster on Jetson Thor and…
Figma revenue jumps 43% to $391M as AI credits start to pay
Figma reported Q3 revenue of $391.4 million, up 43%, with more than 80% of its larger customers consuming AI credits…
Advertisement