Model Follower.

Up to date

Rebuilt
2026-10-26 07:36 UTC
Release radar
116 on watch
Hub snapshot
2026-10-26
Re-checked
daily · every number sourced and dated

Links / Past days

Saturday, 24 October 2026

The AI links the most newsletters and link sources had shared by the end of the day (UTC), each source counted once.

As of 2026-10-24 23:59 UTC Status sealed

That day 36-hour window

  1. Shared by 5

    Kimi K3.1 Pro Tech Blog: Advancing Open Intelligence

    Moonshot releases Kimi K3.1 Pro, a 1-million-token context model with Highspeed at 2-3x speed. Details Terminal-Bench 4.0 scores of 65.7% and xdev-bench-hard vulnerability rates of 81.0%. cited

    Reported byAlphaSignal, Simon Willison, Latent Space / AINews

    Discussed onHacker News 1,060 pts · 654 comments, Techmeme

    AboutKimi-K3

    Where it was shared
    kimi.ai announcement · 13 min read First shared 2026-10-23
  2. Shared by 2

    No Signs of AI Payoff Outside the Tech Sector

    Revenue and profit margins remain flat for the S&P 493 despite massive AI investments in corporate America. Details how mainstream firms view token spending as a cost center and lists specific human factors AI cannot replicate. cited

    Discussed onHacker News 492 pts · 492 comments, Techmeme

    apollo.com analysis · 1 min read First shared 2026-10-24
  3. Shared by 3

    Zagitta/gpu-unlock: Unlock FP/BF16 instruction issue rate on Nvidia 3000's cards to their professional equivalents

    A Linux kernel module doubles FP/BF16 instruction rates on Nvidia 3000 series cards to match professional equivalents. Includes a table showing 3090 Ti gains from 87.46 TFLOPS to 118.39 TFLOPS for FP16 and 87.50 TFLOPS to 157.41 TFLOPS for BF16. cited

    Reported byAlphaSignal, Latent Space / AINews

    Discussed onHacker News 195 pts · 61 comments

    Where it was shared
    github.com repo · 6 min read First shared 2026-10-24
  4. Shared by 1

    packages FLUX.2 [dev] weights for drop-in ComfyUI use, now past 1M downloads.

    Reported byAlphaSignal

    huggingface.co First shared 2026-10-24
  5. Shared by 1

    Diffusion transformer flux2_dev_fp8mixed.safetensors 35.5 GB file

    Reported byAlphaSignal

    huggingface.co First shared 2026-10-24
  6. Shared by 1

    18 GB FP8 file

    “BF16, FP8-mixed, and FP4-mixed variants; the 18 GB FP8 file offers a practical middle ground” via AlphaSignal

    Reported byAlphaSignal

    huggingface.co First shared 2026-10-24
  7. Shared by 1

    AI and the Fall? of the Creative Class

    US creative industries lost over 200k jobs in four years, with movie and sound recording sectors bleeding nearly 100k positions. Compares current job losses to the music industry's streaming transition and details how AI-generated content now competes with traditional video revenue. cited

    Discussed onHacker News 183 pts · 151 comments

    apricitas.io analysis · 10 min read First shared 2026-10-24
  8. Shared by 1

    anthropics margin openais opportunity

    Discussed onTechmeme

    theinformation.com First shared 2026-10-24
  9. Shared by 2

    Push-Based KV Transfer to Improve TTFT and TPOT for Disaggregated Inference in vLLM

    vLLM adds a push-based NIXL connector that moves KV transfer setup off the critical path to improve TTFT and TPOT. Includes benchmarks on Qwen3-32B showing up to 3x better TTFT and 30% better TPOT compared to pull mode at high request rates. cited

    Reported byTLDR AI, AlphaSignal

    Where it was shared
    vllm.ai announcement · 12 min read First shared 2026-10-15
  10. Shared by 5

    deepseek-ai/DeepInfer: DeepInfer: High-performance, large-scale inference framework for DeepSeek models

    DeepSeek releases DeepInfer, an open-source inference framework supporting DeepSeek-V4.1-Flash on NVIDIA SM100 GPUs and Huawei Ascend 950 NPUs. Includes configuration for DSpark verification, SWA bounded replay, and batch-invariant inference, plus build instructions requiring Python 3.12 and CUDA 13. cited

    Reported byInterconnects, Latent Space / AINews, The Rundown AI, AlphaSignal

    Discussed onHacker News 2,346 pts · 891 comments

    Where it was shared
    github.com repo · 2 min read First shared 2026-10-20
  11. Shared by 1

    Serving DeepSeek-V4.1-Flash at Scale: vLLM Matches DeepInfer and Scales to GB300 NVL72

    vLLM serves DeepSeek-V4.1-Flash on an NVIDIA GB300 NVL72 rack at 954K tokens/s per GPU with a P90 of 196 tokens/s per user. Details how Wide-EP scaling and EPLB balancing recover 97% of the gap to uniform routing while overcoming CPU overhead at NVL72 scale. cited

    Reported byAlphaSignal

    AboutDeepSeek-V4.1-Flash

    vllm.ai news report · 11 min read First shared 2026-10-24
  12. Shared by 1

    Perf][EPLB][DSv4] Add a symm_mem weight-migration backend and fuse MegaMoE replica mapping by abmfy · Pull Request #62009

    This PR adds an opt-in symm_mem weight-migration backend and fuses MegaMoE replica mapping into DeepGEMM kernels. Includes per-layer receive buffers in CUDA symmetric memory and async EPLB commits with pinned non-blocking copies. cited

    Reported byAlphaSignal

    github.com repo · 3 min read First shared 2026-10-24
  13. Shared by 1

    vllm-project/vllm: A high-throughput and memory-efficient inference and serving engine for LLMs

    The vLLM project launched a new website at vllm.ai to help users get started with the library. Details supported architectures including DeepSeek-V3 and Mamba, plus installation via uv pip install vllm. cited

    Reported byAlphaSignal

    github.com announcement · 2 min read First shared 2026-10-24
  14. Shared by 1

    alpha school ai education parents.html

    Discussed onTechmeme

    nytimes.com First shared 2026-10-24
  15. Shared by 1

    Unitree’s H2-X Flips Off Cars in New Kung Fu “Guardian” Video

    Unitree released a video showing the H2-X robot performing stunts including a four-meter jump and car flips. Details that the clip shows scripted rescue scenes with a dummy and admits the endurance claim refers only to standing. cited

    Reported bySuperhuman AI

    humanoidsdaily.com news report · 3 min read First shared 2026-10-24
  16. Shared by 2

    Kata Containers, gVisor, Moby/Docker and the Linux kernel

    “Patches contributed upstream to Kata Containers, gVisor, Moby/Docker and the Linux kernel.” via AlphaSignal

    Reported byAlphaSignal

    Discussed onTechmeme

    openai.com First shared 2026-10-23
  17. Shared by 1

    Comfy-Org's repackaged Qwen-Image-Edit bundle for ComfyUI crosses 1.68M downloads.

    Reported byAlphaSignal

    huggingface.co First shared 2026-10-24
  18. Shared by 1

    Qwen-Image-Edit-2511 ComfyUI Native Workflow Example

    Qwen-Image-Edit-2511 adds integrated LoRA capabilities and multi-person consistency to its text rendering editing features. Includes a ComfyUI native workflow example demonstrating material replacement on objects like leather sofas using reference images. cited

    Reported byAlphaSignal

    docs.comfy.org repo · 1 min read First shared 2026-10-24
  19. Shared by 2

    Kimi K3.1 Pro - Intelligence, Performance & Price Analysis

    Kimi K3.1 Pro scores 55 on the Intelligence Index while costing $2.00 per 1M input tokens. Details 2.8 trillion parameters, 1M token context, 77 tokens per second speed, and 150M generated tokens. cited

    Reported bySimon Willison

    Discussed onHacker News 120 pts · 41 comments

    AboutKimi-K3

    artificialanalysis.ai analysis · 4 min read First shared 2026-10-23
  20. Shared by 1

    a veteran of the telecom bust navigates the ai boom b063fa63

    Discussed onTechmeme

    wsj.com First shared 2026-10-24

The week to that day two or more sources

  1. Shared by 5

    Kimi K3.1 Pro Tech Blog: Advancing Open Intelligence

    Moonshot releases Kimi K3.1 Pro, a 1-million-token context model with Highspeed at 2-3x speed. Details Terminal-Bench 4.0 scores of 65.7% and xdev-bench-hard vulnerability rates of 81.0%. cited

    Reported byAlphaSignal, Simon Willison, Latent Space / AINews

    Discussed onHacker News 1,060 pts · 654 comments, Techmeme

    AboutKimi-K3

    Where it was shared
    kimi.ai announcement · 13 min read First shared 2026-10-23
  2. Shared by 6

    Introducing Claude Duet

    Anthropic launches Claude Duet, a proactive agent for Pro and Max users that works on a separate cloud computer. Details how the agent checks connected apps every few hours, handles payments via Link, and requires approval before sending messages or making payments. cited

    Reported byAlphaSignal, The Neuron, The Rundown AI, TLDR AI, Latent Space / AINews, AI Breakfast

    Where it was shared
    claude.com announcement · 7 min read First shared 2026-10-21
  3. Shared by 7

    OpenAI models are coming to Google Cloud

    “OpenAI's frontier models and Codex agentic workflows are coming to Google Cloud's Model Garden.” via AlphaSignal

    Reported byAlphaSignal, The Neuron, The Rundown AI, TLDR AI, Don't Worry About the Vase, AI Breakfast

    Discussed onHacker News 325 pts · 148 comments

    How 1 source put it
    openai.com First shared 2026-10-21
  4. Shared by 6

    OpenAI’s model helps catch critical PHP flaws

    “OpenAI coordinated with PHP maintainers and security partners to patch critical vulnerabilities discovered using GPT-5.6 Sol and independently by Wiz. The company also released a PHP vulnerability checker and proposed a framework for coordinating responses to AI-discovered security flaws.” via TLDR AI

    Reported byAlphaSignal, The Rundown AI, Superhuman AI, TLDR AI, AI Breakfast

    Discussed onHacker News 266 pts · 183 comments

    How 1 source put it
    openai.com First shared 2026-10-21
  5. Shared by 5

    LLMs are making formal software verification feasible at a much larger scale

    Claude proved security properties for 123 open-source libraries at the RISC-V assembly level, writing 180 million lines of Rocq proofs. Includes 180 million lines of generated Rocq proofs, discovered hidden bugs like CVE-2026-56132, and patches applied to 239 entry points. cited

    Reported byAlphaSignal, The Neuron, AI Breakfast

    Discussed onHacker News 457 pts · 305 comments, Techmeme

    Where it was shared
    anthropic.com paper · 15 min read First shared 2026-10-22
  6. Shared by 5

    deepseek-ai/DeepInfer: DeepInfer: High-performance, large-scale inference framework for DeepSeek models

    DeepSeek releases DeepInfer, an open-source inference framework supporting DeepSeek-V4.1-Flash on NVIDIA SM100 GPUs and Huawei Ascend 950 NPUs. Includes configuration for DSpark verification, SWA bounded replay, and batch-invariant inference, plus build instructions requiring Python 3.12 and CUDA 13. cited

    Reported byInterconnects, Latent Space / AINews, The Rundown AI, AlphaSignal

    Discussed onHacker News 2,346 pts · 891 comments

    Where it was shared
    github.com repo · 2 min read First shared 2026-10-20
  7. Shared by 3

    Zagitta/gpu-unlock: Unlock FP/BF16 instruction issue rate on Nvidia 3000's cards to their professional equivalents

    A Linux kernel module doubles FP/BF16 instruction rates on Nvidia 3000 series cards to match professional equivalents. Includes a table showing 3090 Ti gains from 87.46 TFLOPS to 118.39 TFLOPS for FP16 and 87.50 TFLOPS to 157.41 TFLOPS for BF16. cited

    Reported byAlphaSignal, Latent Space / AINews

    Discussed onHacker News 195 pts · 61 comments

    Where it was shared
    github.com repo · 6 min read First shared 2026-10-24
  8. Shared by 3

    Introducing Flint

    Recursive launches Flint, a 40 billion parameter MoE coding model built in under four months. Includes Apache 2.0 weights, rec inference engine, and benchmarks showing 125 tokens per second on an M4 Pro MacBook Pro. cited

    Reported byThe Rundown AI, TLDR AI, Latent Space / AINews

    Where it was shared
    recursive.com announcement · 5 min read First shared 2026-10-22
  9. Shared by 7

    Investing in the public benefit

    Anthropic commits 5% of shares to a new Public Benefit Fund, adding 2% for every 50% valuation increase. Details how the LLC avoids lobbying, funds labor impact studies, and combines with founder pledges to reach 20% total public good equity. cited

    Reported byThe Rundown AI, The Neuron, Superhuman AI, TLDR AI, AI Breakfast, Latent Space / AINews, Forward Future

    Where it was shared
    anthropic.com announcement · 2 min read First shared 2026-10-19
  10. Shared by 2

    OpenAI nears up to $30 billion funding round at $1.25 trillion valuation led by UAE investors

    OpenAI raises up to $30 billion at a $1.25 trillion valuation led by United Arab Emirates investors. Includes a commercial deal to secure more computing capacity from the UAE alongside the new investment. cited

    Reported byThe Rundown AI, TLDR AI

    Where it was shared
    cnbc.com news report · 1 min read First shared 2026-10-23