anthropics/claude-code v2.1.310
Anthropic releases Claude Code version 2.1.310, which adds the Monitor tool to Bedrock, Vertex, and Foundry, and fixes several bugs. ·
New release v2.1.310 of anthropics/claude-code on GitHub.
Top now
6 storiesWhat is climbing fastest across every source
OpenAI6.5xKimi K3.17.1xTool9.9xKimi K3.1 Pro7.0xGoogle8.7xLLMs4.8xChatGPT7.1xClaude5.4xNVIDIA5.5xSpaceX4.8x
Google will pay people to use less electricity so its data centers can use more
Google pays people to reduce electricity use so its data centers can access more power. The company is building a 1GW virtual power plant. ·
anthropics/anthropic-sdk-python v2.3.0
Anthropic releases Python SDK version 2.3.0, adding Claude Tag endpoints for Enterprise organizations and fixing Google Cloud user profile requests. ·
Why Do AI Models Tell Left-Wing Voters to Support the Communist Party?
16 points and 7 comments on Hacker News, linking fsi.stanford.edu.
Quick Study
Quick Study Prompt to Claude Opus 5.5 High builds an artifact that rapidly paints a photo using a canvas element and simulated paintbrush. ·
Turn notifications into code with GitHub Copilot in Teams
GitHub releases a new Microsoft Teams app that combines repository notifications with GitHub Copilot to help teams convert updates into code. ·
anthropics/claude-code v2.1.310
Anthropic releases Claude Code version 2.1.310, which adds the Monitor tool to Bedrock, Vertex, and Foundry, and fixes several bugs. ·
New release v2.1.310 of anthropics/claude-code on GitHub.
Securing our frontier model training environment
61 points and 71 comments on Hacker News, linking openai.com.
Why Do AI Models Tell Left-Wing Voters to Support the Communist Party?
16 points and 7 comments on Hacker News, linking fsi.stanford.edu.
unslothai/unsloth MOSS-Image, Flint & NVIDIA Bolt E4B
Unsloth releases local support for Recursive's Flint, MOSS-Image, and NVIDIA's Nemotron Bolt 4B, alongside a chat-only desktop installer and expanded Mac memory handling. ·
briaai/Fibo-2.0-dev
A text to image model from briaai with 120 likes and 753 downloads on Hugging Face.
MiniMaxAI/MiniMax-M3.1-Flash
A image text to text model from MiniMaxAI with 256 likes and 1,096 downloads on Hugging Face.
Securing our frontier model training environment
61 points and 71 comments on Hacker News, linking openai.com.
briaai/Fibo-2.0-dev
A text to image model from briaai with 120 likes and 753 downloads on Hugging Face.
Dessa: the most authentic full-duplex speech model ever built
Dessa is a research preview of a full-duplex speech model that learns to talk by listening. ·
unslothai/unsloth MOSS-Image, Flint & NVIDIA Bolt E4B
Unsloth releases local support for Recursive's Flint, MOSS-Image, and NVIDIA's Nemotron Bolt 4B, alongside a chat-only desktop installer and expanded Mac memory handling. ·
WeVisOne-VL: Unified Visual Perception from Objects to Pixels
221 upvotes on Hugging Face Daily Papers.
Making MLLM Latents Matter: Factor-Aware Predictive Learning for Cinematic Video Generation
Researchers propose Factor-Aware Predictive Learning to help video diffusion models use multimodal language model latents for cinematic generation, enabling factor-aware control and temporal inference from sparse inputs. ·
Trending topics today: heat
Table
| OpenAI | 2.23 |
| Kimi K3.1 | 1.58 |
| Tool | 1.26 |
| Kimi K3.1 Pro | 1.25 |
| 1.08 | |
| LLMs | 1.03 |
Heat = summed hotness of today's stories naming the topic.
Models
5 storiesReleases, papers, trending weights and new listings
Top 6 trending on Hugging Face: likes
Table
| Lightricks/LTX-2.5 ▲2 | 9033 |
| deepseek-ai/DeepSeek-V4.1-Flash-1017 | 2580 |
| black-forest-labs/flux-3-image-klein-comfy ▼1 | 1405 |
| RecursiveSI/flint ▲2 | 685 |
| OpenMOSS-Team/MOSS-Image-SFT ▼1 | 647 |
| XHToken/Spark-X2.6-4B ▲13 | 410 |
PirateBench: score out of 100
Table
| 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/
Securing our frontier model training environment
61 points and 71 comments on Hacker News, linking openai.com.
Dessa: the most authentic full-duplex speech model ever built
Dessa is a research preview of a full-duplex speech model that learns to talk by listening. ·
briaai/Fibo-2.0-dev
A text to image model from briaai with 120 likes and 753 downloads on Hugging Face.
MiniMaxAI/MiniMax-M3.1-Flash
A image text to text model from MiniMaxAI with 256 likes and 1,096 downloads on Hugging Face.
WeVisOne-VL: Unified Visual Perception from Objects to Pixels
221 upvotes on Hugging Face Daily Papers.
Securing our frontier model training environment
61 points and 71 comments on Hacker News, linking openai.com.
briaai/Fibo-2.0-dev
A text to image model from briaai with 120 likes and 753 downloads on Hugging Face.
Dessa: the most authentic full-duplex speech model ever built
Dessa is a research preview of a full-duplex speech model that learns to talk by listening. ·
WeVisOne-VL: Unified Visual Perception from Objects to Pixels
221 upvotes on Hugging Face Daily Papers.
Making MLLM Latents Matter: Factor-Aware Predictive Learning for Cinematic Video Generation
Researchers propose Factor-Aware Predictive Learning to help video diffusion models use multimodal language model latents for cinematic generation, enabling factor-aware control and temporal inference from sparse inputs. ·
X voices
5 storiesX posts other sources are linking to, then named voices on Bluesky and Mastodon
@DesignArena
Canva's image model reaches #4 on Design Arena
3 points and 0 comments on Hacker News, linking an X post.
@brianchau57
Effective Altruist Hiring Scheme Caused and Covered Up AI Cybersecurity Problems
2 points and 0 comments on Hacker News, linking an X post.
@AgentMemoryL
How Do AI Agents Remember Things Across Tasks?
2 points and 0 comments on Hacker News, linking an X post.
@nickrubin
The gap between GPUs and compute
3 points and 0 comments on Hacker News, linking an X post.
@WeizmanGal
Markdown Is the New Executable: Attackers Drop Agents.md to Your Downloads
4 points and 0 comments on Hacker News, linking an X post.
GitHub
5 storiesNew repos, releases and changelog
New AI repos this week: stars
Table
| aicodeking/leanrea | 1816 |
| alchaincyf/huashu-motion | 1537 |
| wesbos/3d-model-anything | 1065 |
| aicodeking/pandacode | 594 |
| dominikmartn/harp | 591 |
| realskyrin/ns-game-apple-arm-port | 488 |
unslothai/unsloth MOSS-Image, Flint & NVIDIA Bolt E4B
Unsloth releases local support for Recursive's Flint, MOSS-Image, and NVIDIA's Nemotron Bolt 4B, alongside a chat-only desktop installer and expanded Mac memory handling. ·
New release v0.1.910-beta of unslothai/unsloth on GitHub.
ollama/ollama v0.40.10
Ollama releases version 0.40.10, adding support for NVIDIA's Nemotron 3.5 Bolt E4B, a Gemma 4-based vision-language model that supports a 128K-token context window. ·
openai/codex 0.164.0
OpenAI releases Codex 0.164.0, which adds ten preset slots for prompts, inline model arguments, and image previews. The update also enables the Amazon Bedrock setup wizard by default. ·
anthropics/claude-code v2.1.309
Anthropic releases Claude Code version 2.1.309, which fixes sandboxing, permission checks, and worktree isolation issues on macOS and Linux. ·
OpenShell – A safe, private runtime for fleets of autonomous AI agents
NVIDIA introduces OpenShell, a safe and private runtime designed for fleets of autonomous AI agents. ·
unslothai/unsloth MOSS-Image, Flint & NVIDIA Bolt E4B
Unsloth releases local support for Recursive's Flint, MOSS-Image, and NVIDIA's Nemotron Bolt 4B, alongside a chat-only desktop installer and expanded Mac memory handling. ·
New release v0.1.910-beta of unslothai/unsloth on GitHub.
ollama/ollama v0.40.10
Ollama releases version 0.40.10, adding support for NVIDIA's Nemotron 3.5 Bolt E4B, a Gemma 4-based vision-language model that supports a 128K-token context window. ·
anthropics/claude-code v2.1.309
Anthropic releases Claude Code version 2.1.309, which fixes sandboxing, permission checks, and worktree isolation issues on macOS and Linux. ·
OpenShell – A safe, private runtime for fleets of autonomous AI agents
NVIDIA introduces OpenShell, a safe and private runtime designed for fleets of autonomous AI agents. ·
openai/codex 0.164.0
OpenAI releases Codex 0.164.0, which adds ten preset slots for prompts, inline model arguments, and image previews. The update also enables the Amazon Bedrock setup wizard by default. ·
Infrastructure
5 storiesData centers, power, chips and what tokens cost
Data-center and energy stories per day
Table
| Sun 18 | 2 |
| Mon 19 | 32 |
| Tue 20 | 31 |
| Wed 21 | 29 |
| Thu 22 | 34 |
| Fri 23 | 39 |
| Sat 24 | 0 |
Stories from DCD, Data Center Knowledge, EIA and Hacker News, by publication day.
Median output price of models added in the last 30 days, $ per million tokens
One point per OpenRouter reading, last 8 days. Data by OpenRouter, CC BY 4.0.
Output price, $ per million tokens (newest models)
Table
| kimi-k3.1-pro | 10 |
| minimax-m3.1-flash | 0.6 |
| deepseek-v4.1-flash-1017 | 0.6 |
| glimpse | 3 |
| beam-1.0-501b-a23b | 1 |
| median, all paid models | 2 |
440 paid models listed; median of the 31 added in 30 days: $3.20. Rankings data by OpenRouter, CC BY 4.0.
Data Center World Middle East
From Data Center Knowledge.
Crown Castle bullish on carrier spectrum deployment, says it can support SpaceX's terrestrial dreams
Crown Castle expresses confidence in carrier spectrum deployment and states it can support SpaceX's terrestrial rollout plans. ·
Google open-sources TorchTPU, a native PyTorch back end for TPUs
379 points and 90 comments on Hacker News, linking github.com.
Texas data center boom could erase decades of progress on air quality
A new report warns that Texas data center expansion could reverse the state's air quality progress by the end of the decade, particularly if facilities rely on natural gas power. ·
Data Center World Middle East
From Data Center Knowledge.