Archived front page, Tue 29 Sep 2026, 11:00 UTC. All of 2026-09-29 · Archive · Live page

The Daily Markdown

The front page of the internet for AI users · Tue 29 Sep 2026, 11:00 UTC

Top now

6 stories

What is climbing fastest across every source

OpenAI6.4xWorld Labs7.4xLearning6.6xAnthropic6.9xClaude Sonnet 5.57.0xGeneral Agents6.8xOpus 5.57.0xAMD7.4xCorrection6.6xNvidia6.1x

openai/codex 0.159.0

OpenAI released version 0.159.0 of Codex, adding an opt-in feature that allows new input to steer the model during active responses or long-running code calls. ·

GitHub Releases3h ago▲ 23github.com

openai/codex 0.159.0

OpenAI released version 0.159.0 of Codex, adding an opt-in feature that allows new input to steer the model during active responses or long-running code calls. ·

GitHub Releases3h ago▲ 89github.com

Claude Sonnet 5.5

191 points and 138 comments on Hacker News, linking anthropic.com.

Hacker News+ Simon Willison17h ago▲ 7665 to 329 across 4 pollsdiscussanthropic.com

Post-Training Leaves Behavioral Shadows on Unrelated Decisions

Researchers introduced Active Taskless Distillation, a method transferring capabilities using single-word prompts. In coding tests, this approach yielded a 5.34 percentage point gain on HumanEval+. ·

HF Daily Papers11h ago▲ 6311 to 70 across 5 pollsarxiv.org

KKKKhazix/AIHOT

A website framework automatically identifies trending topics and generates daily reports. Users can customize its sources and selection criteria to create industry-specific news sites. ·

GitHub11h ago▲ 62462 to 2292 across 10 pollsgithub.com

Claude Sonnet 5.5

191 points and 138 comments on Hacker News, linking anthropic.com.

Hacker News+ Simon Willison17h ago▲ 12265 to 329 across 4 pollsdiscussanthropic.com

JoasASantos/Offensive-Security-AI-Models

JoasASantos maintains a repository of AI models that are uncensored or fine-tuned for cybersecurity tasks. ·

GitHub+ Hacker News38h ago▲ 119110 to 291 across 20 pollsgithub.com

openai/codex 0.159.0

OpenAI released version 0.159.0 of Codex, adding an opt-in feature that allows new input to steer the model during active responses or long-running code calls. ·

GitHub Releases3h ago▲ 96github.com

Post-Training Leaves Behavioral Shadows on Unrelated Decisions

Researchers introduced Active Taskless Distillation, a method transferring capabilities using single-word prompts. In coding tests, this approach yielded a 5.34 percentage point gain on HumanEval+. ·

HF Daily Papers11h ago▲ 8611 to 70 across 5 pollsarxiv.org

KKKKhazix/AIHOT

A website framework automatically identifies trending topics and generates daily reports. Users can customize its sources and selection criteria to create industry-specific news sites. ·

GitHub11h ago▲ 85462 to 2292 across 10 pollsgithub.com

Trending topics today: heat

OpenAI: 4.01OpenAI4.01World Labs: 1.79World Labs1.79Learning: 1.26Learning1.26Anthropic: 1.07Anthropic1.07Claude Sonnet 5.5: 1.06Claude Sonnet 5.51.06General Agents: 1.05General Agents1.05
Table
OpenAI4.01
World Labs1.79
Learning1.26
Anthropic1.07
Claude Sonnet 5.51.06
General Agents1.05

Heat = summed hotness of today's stories naming the topic.

Models

5 stories

Releases, papers, trending weights and new listings

Top 6 trending on Hugging Face: likes

convaiinnovations/laya: 4,415convaiinnovations/laya4,415Qwen/Qwen-Image-2.1: 2,624Qwen/Qwen-Image-2.12,624abenzerps/Qwen-Image-2.1-Uncensored-GGUF: 2,338abenzerps/Qwen-Image-…2,338XingChen-AGI/Xing4.0-29B-A4B: 1,803XingChen-AGI/Xing4.0-…1,803Edge0/Audio8-ASR-Infinite: 1,458Edge0/Audio8-ASR-Infi…1,458XingChen-AGI/TeleOCR: 830XingChen-AGI/TeleOCR830
Table
convaiinnovations/laya4415
Qwen/Qwen-Image-2.12624
abenzerps/Qwen-Image-2.1-Uncensored-GGUF2338
XingChen-AGI/Xing4.0-29B-A4B1803
Edge0/Audio8-ASR-Infinite1458
XingChen-AGI/TeleOCR830

PirateBench: score out of 100

opus-5-5: 91opus-5-591gpt-6-sol: 82gpt-6-sol82deepseek-v4-flash-0731: 63deepseek-v4-flash-073163
Table
opus-5-591.15
gpt-6-sol81.78
deepseek-v4-flash-073162.82

Agents build our own projects from a brief; every run, failures included: https://cowerx.dev/daily/piratebench/

Post-Training Leaves Behavioral Shadows on Unrelated Decisions

Researchers introduced Active Taskless Distillation, a method transferring capabilities using single-word prompts. In coding tests, this approach yielded a 5.34 percentage point gain on HumanEval+. ·

HF Daily Papers11h ago▲ 011 to 70 across 5 pollsarxiv.org

Claude Sonnet 5.5

191 points and 138 comments on Hacker News, linking anthropic.com.

Hacker News+ Simon Willison17h ago▲ 7665 to 329 across 4 pollsdiscussanthropic.com

Post-Training Leaves Behavioral Shadows on Unrelated Decisions

Researchers introduced Active Taskless Distillation, a method transferring capabilities using single-word prompts. In coding tests, this approach yielded a 5.34 percentage point gain on HumanEval+. ·

HF Daily Papers11h ago▲ 6311 to 70 across 5 pollsarxiv.org

Claude Sonnet 5.5

191 points and 138 comments on Hacker News, linking anthropic.com.

Hacker News+ Simon Willison17h ago▲ 12265 to 329 across 4 pollsdiscussanthropic.com

JoasASantos/Offensive-Security-AI-Models

JoasASantos maintains a repository of AI models that are uncensored or fine-tuned for cybersecurity tasks. ·

GitHub+ Hacker News38h ago▲ 119110 to 291 across 20 pollsgithub.com

Post-Training Leaves Behavioral Shadows on Unrelated Decisions

Researchers introduced Active Taskless Distillation, a method transferring capabilities using single-word prompts. In coding tests, this approach yielded a 5.34 percentage point gain on HumanEval+. ·

HF Daily Papers11h ago▲ 8611 to 70 across 5 pollsarxiv.org

X voices

5 stories

X posts other sources are linking to, then named voices on Bluesky and Mastodon

GitHub

5 stories

New repos, releases and changelog

New AI repos this week: stars

KKKKhazix/AIHOT: 2,292KKKKhazix/AIHOT2,292yetone/magpie: 1,830yetone/magpie1,830dzhng/jevgrep: 1,588dzhng/jevgrep1,588mikehasa/golive-skill: 1,081mikehasa/golive-skill1,081ollaya-dev/ollaya: 927ollaya-dev/ollaya927feitangyuan/onetake: 914feitangyuan/onetake914
Table
KKKKhazix/AIHOT2292
yetone/magpie1830
dzhng/jevgrep1588
mikehasa/golive-skill1081
ollaya-dev/ollaya927
feitangyuan/onetake914

KKKKhazix/AIHOT

A website framework automatically identifies trending topics and generates daily reports. Users can customize its sources and selection criteria to create industry-specific news sites. ·

A new repository written in TypeScript with 2,292 stars in its first week.

GitHub11h ago▲ 0462 to 2292 across 10 pollsgithub.com

langchain-fireworks==1.7.0

LangChain released version 1.7.0 of its Fireworks integration. The update adds prompt caching middleware and fixes mid-stream read timeout classification. ·

GitHub Releases14h ago▲ 0github.com

anthropics/claude-code v2.1.284

Anthropic released Claude Code v2.1.284, making Claude Sonnet 5.5 the default model with a 1M token context window. ·

GitHub Releases17h ago▲ 030 to 49 across 8 pollsgithub.com

KKKKhazix/AIHOT

A website framework automatically identifies trending topics and generates daily reports. Users can customize its sources and selection criteria to create industry-specific news sites. ·

A new repository written in TypeScript with 2,292 stars in its first week.

GitHub11h ago▲ 62462 to 2292 across 10 pollsgithub.com

anthropics/claude-code v2.1.284

Anthropic released Claude Code v2.1.284, making Claude Sonnet 5.5 the default model with a 1M token context window. ·

GitHub Releases17h ago▲ 4830 to 49 across 8 pollsgithub.com

JoasASantos/Offensive-Security-AI-Models

JoasASantos maintains a repository of AI models that are uncensored or fine-tuned for cybersecurity tasks. ·

GitHub+ Hacker News38h ago▲ 41110 to 291 across 20 pollsgithub.com

JoasASantos/Offensive-Security-AI-Models

JoasASantos maintains a repository of AI models that are uncensored or fine-tuned for cybersecurity tasks. ·

A new repository with 291 stars in its first week.

GitHub+ Hacker News38h ago▲ 119110 to 291 across 20 pollsgithub.com

KKKKhazix/AIHOT

A website framework automatically identifies trending topics and generates daily reports. Users can customize its sources and selection criteria to create industry-specific news sites. ·

GitHub11h ago▲ 85462 to 2292 across 10 pollsgithub.com

anthropics/claude-code v2.1.284

Anthropic released Claude Code v2.1.284, making Claude Sonnet 5.5 the default model with a 1M token context window. ·

GitHub Releases17h ago▲ 7730 to 49 across 8 pollsgithub.com

unslothai/unsloth Laya Decision Models + Library

Unsloth added decision model support, a document library, and Apple Silicon optimizations. The update also enables local Laya execution and accelerates image and video generation by up to 4.5 times. ·

GitHub Releases20h ago▲ 683 to 15 across 10 pollsgithub.com

Data centers & energy

5 stories

Builds, power deals and the grid

Data-center and energy stories per day

Wed 23: 3Wed 233Thu 24: 3Thu 243Fri 25: 20Fri 2520Sat 26: 1Sat 261Sun 27: 1Sun 271Mon 28: 20Mon 2820Tue 29: 5Tue 295
Table
Wed 233
Thu 243
Fri 2520
Sat 261
Sun 271
Mon 2820
Tue 295

Stories from DCD, Data Center Knowledge, EIA and Hacker News, by publication day.

Investing for a high performance future

Waterless two-phase liquid cooling supports high-density computing and sustained performance in low-latency trading environments. ·

DCD3h ago▲ 7datacenterdynamics.com

Investing for a high performance future

Waterless two-phase liquid cooling supports high-density computing and sustained performance in low-latency trading environments. ·

DCD3h ago▲ 27datacenterdynamics.com

Investing for a high performance future

Waterless two-phase liquid cooling supports high-density computing and sustained performance in low-latency trading environments. ·

DCD3h ago▲ 29datacenterdynamics.com

Hardware prices

2 stories

GPUs, memory, and what tokens cost

Output price, $ per million tokens (newest models)

claude-sonnet-5.5: 10claude-sonnet-5.510claude-sonnet-5.5:batch: 5claude-sonnet-5.5:bat…5perceptron-mk1.5: 1.50perceptron-mk1.51.50ember-1: 15ember-115glm-5.3-prime: 8.80glm-5.3-prime8.80median, all paid models: 2median, all paid mode…2
Table
claude-sonnet-5.510
claude-sonnet-5.5:batch5
perceptron-mk1.51.5
ember-115
glm-5.3-prime8.8
median, all paid models2

434 paid models listed; median of the 56 added in 30 days: $4.00. Rankings data by OpenRouter, CC BY 4.0.