The front page of the internet for AI users · Fri 16 Oct 2026, 02:00 UTC · Get it by email
What is climbing fastest across every source
Intern Lumina U2 is a diffusion language model that represents multi-codebook visual tokens in a fully discrete space to unify visual understanding and image generation. ·
5 upvotes on Hugging Face Daily Papers.
SWE-at-Scale is a framework that automatically synthesizes realistic software engineering tasks by reverse-engineering historical bug fixes to create diverse evaluation scenarios for coding agents. ·
PointWormGazebo offers a universal cheat panel that provides AI-enhanced aimbot and per-game precision tuning for various video games. ·
From Hugging Face blog.
New research from Bold Agent Kit indicates that human receptionists achieve roughly four times higher inbound intake conversion rates than AI, despite AI handling a larger volume of calls. ·
OpenRouter adds Moondream 3 Preview, a vision-language model with nine billion total parameters and two billion active parameters that handles visual question answering, captioning, OCR, and visual reasoning. ·
PointWormGazebo offers a universal cheat panel that provides AI-enhanced aimbot and per-game precision tuning for various video games. ·
A new repository with 122 stars in its first week.
Intern Lumina U2 is a diffusion language model that represents multi-codebook visual tokens in a fully discrete space to unify visual understanding and image generation. ·
New research from Bold Agent Kit indicates that human receptionists achieve roughly four times higher inbound intake conversion rates than AI, despite AI handling a larger volume of calls. ·
SWE-at-Scale is a framework that automatically synthesizes realistic software engineering tasks by reverse-engineering historical bug fixes to create diverse evaluation scenarios for coding agents. ·
New release v2.1.301 of anthropics/claude-code on GitHub.
A image text to text model from meta-models with 260 likes and 402 downloads on Hugging Face.
PointWormGazebo offers a universal cheat panel that provides AI-enhanced aimbot and per-game precision tuning for various video games. ·
A new repository with 122 stars in its first week.
New research from Bold Agent Kit indicates that human receptionists achieve roughly four times higher inbound intake conversion rates than AI, despite AI handling a larger volume of calls. ·
New release v2.1.301 of anthropics/claude-code on GitHub.
A image text to text model from meta-models with 260 likes and 402 downloads on Hugging Face.
llama.cpp v0.7.0 adds LiliCorr speculative decoding, a GPU cache for host-memory experts, and support for models like K2 Horizon and Prism Bonsai 2. ·
Intern Lumina U2 is a diffusion language model that represents multi-codebook visual tokens in a fully discrete space to unify visual understanding and image generation. ·
| Anthropic | 2.84 |
| black-forest-labs | 1.4 |
| flux-3 | 1.4 |
| Chinese | 1.23 |
| Scion | 1.08 |
| OpenRouter | 0.92 |
Heat = summed hotness of today's stories naming the topic.
Releases, papers, trending weights and new listings
| abenzerps/Qwen-Image-2.1-Uncensored-Quantized | 4390 |
| google/embeddinggemma-2 | 2408 |
| jialinyyzz/humanizer | 1738 |
| Qwen/Qwen-Image-2.1-Turbo | 881 |
| openbmb/MiniCPM-V-4.7-35B-A3B | 794 |
| fishaudio/drama-3 ▲5 | 479 |
| gpt-6.1-sol | 98.9 |
| opus-5-5 | 96.46 |
| gpt-6-sol | 91.6 |
| gpt-6-luna | 65.25 |
| haiku-4-5 | 4.28 |
Agents build our own projects from a brief; every run, failures included: https://cowerx.dev/daily/piratebench/
A image text to text model from meta-models with 260 likes and 402 downloads on Hugging Face.
A image to image model from black-forest-labs with 222 likes and 39 downloads on Hugging Face.
A image to image model from black-forest-labs with 234 likes and 3,241 downloads on Hugging Face.
Krea 3 is a new image editing model designed for production use. It preserves product integrity and offers volume-based pricing. ·
The Simons Foundation launches a $15 million effort to help mathematicians navigate the rapid rise of artificial intelligence in their field. ·
A image text to text model from meta-models with 260 likes and 402 downloads on Hugging Face.
A image to image model from black-forest-labs with 234 likes and 3,241 downloads on Hugging Face.
A image to image model from black-forest-labs with 222 likes and 39 downloads on Hugging Face.
Krea 3 is a new image editing model designed for production use. It preserves product integrity and offers volume-based pricing. ·
The Simons Foundation launches a $15 million effort to help mathematicians navigate the rapid rise of artificial intelligence in their field. ·
X posts other sources are linking to, then named voices on Bluesky and Mastodon
Handwired: Turn your regular JavaScript code into a neural net
4 points and 0 comments on Hacker News, linking an X post.
Total War Remaster Removed from Steam After AI Allegations, Brings Back Original
4 points and 0 comments on Hacker News, linking an X post.
DeepSeek v4.1 Flash is now free in OpenCode
4 points and 0 comments on Hacker News, linking an X post.
AI researchers are taking up smoking to signal to other researchers
3 points and 0 comments on Hacker News, linking an X post.
Can Coding Agents Act Like Staff Engineers?
2 points and 1 comments on Hacker News, linking an X post.
Handwired: Turn your regular JavaScript code into a neural net
4 points and 0 comments on Hacker News, linking an X post.
Total War Remaster Removed from Steam After AI Allegations, Brings Back Original
4 points and 0 comments on Hacker News, linking an X post.
DeepSeek v4.1 Flash is now free in OpenCode
4 points and 0 comments on Hacker News, linking an X post.
Can Coding Agents Act Like Staff Engineers?
2 points and 1 comments on Hacker News, linking an X post.
OpenAI finds recording device in one of their offices
85 points and 2 comments on Hacker News, linking an X post.
New repos, releases and changelog
| onehumanbeing/esp32-iphone-agent | 2492 |
| rociiu/talorys | 1159 |
| marclou/mailcheap | 1055 |
| zhuyansen/new-term-radar | 1026 |
| deeplethe/solid-glass | 1016 |
| outcrawlai/scout | 832 |
New release v2.1.301 of anthropics/claude-code on GitHub.
Unsloth adds local support for NVIDIA Nemotron 3.5 and Meta Muse Spark Open, introducing text-diffusion LLMs, DeepSpeed training, and improved prompt caching for coding agents. ·
llama.cpp v0.7.0 adds LiliCorr speculative decoding, a GPU cache for host-memory experts, and support for models like K2 Horizon and Prism Bonsai 2. ·
Feder-cr releases open_scion, an open-source version control system designed specifically for coding agents. ·
BlastLoop is a structure-first coding method for AI agents that uses build altitudes and quality gates to manage development. ·
New release v2.1.301 of anthropics/claude-code on GitHub.
llama.cpp v0.7.0 adds LiliCorr speculative decoding, a GPU cache for host-memory experts, and support for models like K2 Horizon and Prism Bonsai 2. ·
Feder-cr releases open_scion, an open-source version control system designed specifically for coding agents. ·
Unsloth adds local support for NVIDIA Nemotron 3.5 and Meta Muse Spark Open, introducing text-diffusion LLMs, DeepSpeed training, and improved prompt caching for coding agents. ·
New release v2.1.300 of anthropics/claude-code on GitHub.
Data centers, power, chips and what tokens cost
| Sat 10 | 3 |
| Sun 11 | 7 |
| Mon 12 | 41 |
| Tue 13 | 29 |
| Wed 14 | 34 |
| Thu 15 | 41 |
| Fri 16 | 0 |
Stories from DCD, Data Center Knowledge, EIA and Hacker News, by publication day.
One point per OpenRouter reading, last 8 days. Data by OpenRouter, CC BY 4.0.
| moondream3-preview | 2.5 |
| moondream3.1-9b-a2b | 2.5 |
| muse-spark-open | 4.25 |
| nemotron-3.5-super-vl | 0.5 |
| pareto-latest | 7.5 |
| median, all paid models | 2 |
438 paid models listed; median of the 47 added in 30 days: $5.00. Rankings data by OpenRouter, CC BY 4.0.
49 points and 40 comments on Hacker News, linking reuters.com.
Proposed reliability standards for large AI data centers and crypto miners await approval by NERC and federal regulators, applying to loads of 50 MW or more at 100 kV or higher. ·
378 points and 272 comments on Hacker News, linking startupfortune.com.
49 points and 40 comments on Hacker News, linking reuters.com.
378 points and 272 comments on Hacker News, linking startupfortune.com.
60 points and 39 comments on Hacker News, linking dfarq.homeip.net.
25 points and 39 comments on Hacker News, linking reuters.com.
Small AI models may soon self-replicate across cheap IoT devices like Android TV boxes without requiring a datacenter. ·
