Hackernews posts about MLX

1. Qwen3.8-27B-Uncensored-MLX (huggingface.co)
   7 points by worldofmatthew 3 days ago | 2 comments
2. Show HN: mlxsh A lightweight CLI to serve multiple local LLMs on Apple (github.com)
   3 points by ansnadeem 20 days ago | 1 comments
3. The Qwen MLX Challenge (www.yukon.org)
   2 points by vivekkairi 6 days ago | discuss
4. From CUDA to MLX: How K-Search Brings Kernel Expertise to Apple Silicon (bair.berkeley.edu)
   2 points by williamjinq 23 days ago | discuss
5. Explore distributed inference and training with MLX [video] (developer.apple.com)
   2 points by pulkitsh1234 30 days ago | discuss
6. Apertura A from-scratch Objective-C++ / MLX rebuild of Gemma-4 for Apple Silicon (github.com)
   1 points by Bluestein 4 days ago | 1 comments
7. Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
   88 points by louis030195 30 days ago | 67 comments
8. Show HN: I shrank DeepSeek V4 Flash to 57GB and it wrote a compiler on my Mac (huggingface.co)
   21 points by hacklas 6 days ago | 3 comments
9. Show HN: Local text, image, video, music and 3D from one CLI, no Python (github.com)
   16 points by sawfwair 23 days ago | 7 comments
10. Show HN: Writekin – fine-tune a local LLM on your own writing, on your Mac (github.com)
    6 points by eggbrain 25 days ago | 3 comments
11. Show HN: Minna – On-device semantic search across your own documents (macOS) (github.com)
    3 points by ActuallyTaylor 3 days ago | discuss
12. Qwen3.8-Max: A New Bar for Coding and Cowork (qwen.ai)
    1124 points by ai2027 20 days ago | 615 comments
13. Qwen3.8 Max now ranked as the best overall model by agentic index (artificialanalysis.ai)
    546 points by apitman 16 days ago | 354 comments
14. Maximizing the value of your Claude Code sessions (claude.com)
    319 points by twapi 8 days ago | 184 comments
15. The Maxwell Conjecture Is False (GPT 5.6 Sol) (arxiv.org)
    157 points by rahen 22 days ago | 142 comments
16. Running Kimi K3 on a M1 Max (github.com)
    138 points by tito 25 days ago | 93 comments
17. Show HN: Check if any of the $656M in unclaimed royalties at The MLC is yours (pub.doub.ly)
    76 points by knaught 2 days ago | 37 comments
18. Termixer (TUI DJ Mixer) (github.com)
    68 points by l00sed 22 days ago | 40 comments
19. Code Obfuscation via Local Mixing (vitalik.eth.limo)
    34 points by fbrusch 1 day ago | 1 comments
20. Show HN: Max Studio Tools – C++ DSP Modules for Max and Ableton Live (github.com)
    32 points by apresta 29 days ago | discuss
21. I built a f.lux alternative because the M5 Max firmware broke existing solutions (driftformac.app)
    25 points by jackvanstone 22 days ago | 21 comments
22. Qwen 3.8 Max Live Now (www.qwencloud.com)
    25 points by zkmon 20 days ago | 5 comments
23. Auto mode is now the default in Claude Code for Pro, Max, and Team plans (simonwillison.net)
    17 points by spenvo 14 days ago | 5 comments
24. Why Substack Is Collapsing: Subscription Model Has Maxed Out (sgcarney.substack.com)
    12 points by bookofjoe 19 days ago | 7 comments
25. Mixture-of-Kittens: our open-source MoE megakernel for NVL72s (cursor.com)
    12 points by mfiguiere 18 days ago | discuss
26. Show HN: Listnr – macOS meeting transcription that never mixes mic and speakers (github.com)
    11 points by rokib16x 18 days ago | 4 comments
27. Qwen 3.8-Max Preview (manish.sh)
    11 points by ms7892 15 days ago | discuss
28. Boeing 737 MAX operators told to check fuselages for cracks (www.theguardian.com)
    9 points by frereubu 15 days ago | discuss
29. Antirez - streaming the Kimi K3 official hugging face weights on M5 max 128gb (twitter.com)
    8 points by mihau 25 days ago | 1 comments
30. Qwen3.8-Max (twitter.com)
    8 points by logickkk1 20 days ago | discuss