updated 63 minutes ago
Category: All Coding agents MCP AI video Local AI Developer tools Self-hosted Models Skills
Sort: Default order Stars gained, 7d Stars gained, 24h HN points
Language: All C C++ Cuda Go JavaScript Jupyter Notebook Kotlin Python Rust TypeScript
Topic: All ai ai-agent ai-agents apple-silicon claude-code cli cuda electron gguf inference linux llama-cpp llm llm-inference local-ai local-first local-llm macos mcp ollama open-source openai privacy pytorch quantization qwen rust speculative-decoding typescript windows
Marvel Rivals UE5 External Bounding-Box ESP & Ult Tracker: native 64-bit Windows 10/11 por...
MID 42 45 -3 - WatchingDiscord & Local Uncensored AI Deep-Live-Cam v2.1 1-Click Realtime Webcam FaceSwap CUDA Por...
MID 42 45 -3 - WatchingDiscord & Local Uncensored AI Discord Activity serial-utility, Quest Auto-Completer & Badg...
MID 42 46 -4 - WatchingDeep-Live-Cam v2.1 Portable (1-Click CUDA Installer) — Realtime Webcam FaceSwap for OBS/Di...
MID 42 31 +11 - WatchingA quantization compiler that delivers 12× lower KL than Q4_K_M at the same file size
TOO EARLY 99 39 +29 to +71 - Being talked aboutKobiMusic's multi-instrument audio-to-MIDI transcriber: notes, instrument tracks and drums...
MID 42 57 -15 +55 WatchingCalibrated internal-state modulation, readout, and neural observability for local dense la...
MID 42 24 +18 +47 WatchingAI assistant and MCP server for Blender: visual chat, scene tools, multi-agent workflows....
MID 42 19 +23 +813 SurgingPaste a link, get captioned 9:16 clips in ~2 min. Free, local, open-source clipping engine...
MID 42 34 +9 +430 SurgingA desktop app for the Google Colab CLI: accounts, GPU sessions with cost guards, port forw...
SLEEPER 42 75 -33 +36 WatchingThe open LLM inference engine for AWS Trainium: NKI megakernels, 1M-token context, OpenAI-...
OVERHYPED 97 49 +48 +2 Being talked aboutTerminal-native AI coding assistant for real repositories. Inspect code, edit files, run t...
REAL 96 83 +14 +3 WatchingSee inside any GGUF: where two quantizations part ways, token by token, and what each samp...
MID 42 36 +7 +8 WatchingAgent skill for hill-climbing local LLM serving speed (decode, speculative decoding, prefi...
MID 42 17 +26 +391 SurgingLive, multi-camera object tracking with SAM3, SAM 2 and EdgeTAM, on one GPU or many. One m...
OVERHYPED 100 47 +53 +6 Being talked aboutOffline knowledge AI, maps and videos for survival and rebuilding
SLEEPER 42 76 -34 +114 WatchingOpen-source browser extension that flags AI-generated images and video as you browse. Runs...
MID 54 27 +26 +5 WatchingDeepSeek V3 (MLA, DeepSeekMoE, aux-loss-free balancing, MTP) in ~200 lines of pure JAX, tr...
TOO EARLY 94 36 +25 to +67 +1 WatchingAI Manus is a general-purpose AI Agent system that supports running various tools and oper...
SLEEPER 14 70 -56 -4 SlippingA multi-provider AI copilot hub for Blender—bring OpenAI, Anthropic Claude, Google Gemini,...
OVERHYPED 64 15 +49 -1 WatchingLinger is An Open-Source Personal Agent That Runs Fully on Your PC. We aim to break Big Te...
REAL 78 84 -6 +494 SurgingHaru: a low-latency, multilingual, full-duplex spoken dialogue model
TOO EARLY 73 30 +9 to +50 +14 Quiet climberFaster speech off signals for spec gen voice pipelines
OVERHYPED 88 51 +37 0 WatchingNative Windows 11 ARM64 on iOS (iPhone Duo) with direct ARM64 JIT, 4KB ASID MMU, UEFI/ACPI...
TOO EARLY 69 17 +13 to +54 +8 WatchingWorldForge 世界工坊 — Squeeze 100% out of Google Colab GPUs (RTX PRO 6000 Blackwell 98GB VRAM...
MID 29 14 +15 0 Watchingmaxxing the ability to run different LLMs on the smallest devices possible
TOO EARLY 97 66 +17 to +54 +23 Quiet climberSOTA model-agnostic quantization for GGUF, local inference and beyond. Quantize and run te...
TOO EARLY 94 31 +29 to +70 +4 WatchingCompress any LLM in seconds, load it for inference and in one click you have an AI that co...
TOO EARLY 83 26 +20 to +62 0 WatchingPrivate AI meeting notes that stay on your computer. No meeting bot. Free and open source...
REAL 99 88 +11 +89 SurgingTernary language models on a home computer: run, train, convert, fine-tune, export to GGUF
OVERHYPED 90 39 +51 0 WatchingQwen3.5 / 3.8 to ternary with TTF-trainable ternary adapters
TOO EARLY 82 34 +14 to +56 0 WatchingTwo-timescale threshold flips (TTF): training and converting ternary {-1,0,1} LLMs without...
MID 85 59 +25 +3 WatchingNaiveRT: a self-contained TP8 inference runtime for the Naive N series on eight Blackwell...
TOO EARLY 57 0 +17 to +52 +5 WatchingThe fastest inference engine for Qwen3.8-27B on the NVIDIA RTX 5090: up to 500 tokens/s fo...
MID 23 36 -13 0 Watchingsmall LLM inference engine from scratch, w/ 6+ hours of youtube tutorials & 85+ pages of h...
TOO EARLY 92 20 +11 WatchingMAPLE: Manifold-Activated Pruning of Latent Experts for Sub-2-BPW MoE LLMs
TOO EARLY 90 51 +16 to +56 0 WatchingZero-GPU CPU inference for 30B-284B MoE models (Qwen3.8 176.5B cold-cache certified in 28....
REAL 95 70 +25 +5 WatchingHyperFlow H3 Flash: Video Rebirth's 13.7B, 4- to 8-step compressed MiniMax-H3 for joint vi...
TOO EARLY 91 32 +25 to +67 +1 WatchingA GPU task manager for Windows, Linux, and WSL: per-process VRAM, GPU activity, and LLM mo...
REAL 74 81 -8 +32 Quiet climberQwen3.8-Flash-Next-FP8 (48-layer MoE, 512 experts) on one RTX 3070 Ti plus 24 Maxwell GPUs...
MID 22 24 -2 0 WatchingCage:为 Claude 特化的轻量化虚拟机,彻底杜绝网络泄露。多平台适配,支持多账号管理、远程部署、移动端操控。
REAL 77 92 -15 +320 SurgingOpenAmika: MekaHime's AI companion Amika, open source. Live2D avatar, real-time voice, lon...
REAL 59 74 -15 0 WatchingAssistente de sobrevivência 100% offline para Windows, Linux e macOS: guias com fontes (ág...
TOO EARLY 81 23 +20 to +62 0 WatchingEXTREMELY autonomous AI agents: a free, open-source AI agency that runs on your own comput...
SLEEPER 22 72 -50 0 WatchingOpen-source AI coding agent and agent runtime: one binary, any model, MCP-native. Runs in...
REAL 82 93 -11 +3 SlippingQwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata...
MID 22 50 -28 0 WatchingTiny local-first AI harness: any model, MCP and Agent Skills, natural-language CLI. Zero d...
SLEEPER 13 49 -36 0 SlippingComplete logs of the experiments in the blog post 'LLMs can understand and write neural ne...
TOO EARLY 92 7 +1 WatchingThe lossless, headless coding agent harness. One 7 MB static binary, four tools, plain std...
SLEEPER 8 65 -56 +2 SlippingA phone line over USB-C: LTE data, SMS and voice calls with the call audio on USB. A 4-lay...
TOO EARLY 83 36 +13 to +56 +1 WatchingFree local-first Windows companion for job search and everyday work, with editable respons...
MID 42 16 +26 0 SlippingFree local-first AI business advisor with evidence-linked research and verifiable pilot pl...
MID 42 33 +9 0 SlippingA chatbot, with a large language model split across its users, entirely in the browser.
MID 23 51 -27 0 WatchingA 4B decision model: calibrated probabilities for yes/no, choice and score questions over...
MID 42 37 +5 0 SlippingAn on-device AI agent for Android that operates apps and the web to get things done for yo...
REAL 76 97 -21 +201 Quiet climberTiny world model generates games and stream ac plays them all on Pico 2W
TOO EARLY 87 19 0 WatchingA SIP phone daemon with local speech and language models on Apple Silicon.
TOO EARLY 84 52 +9 to +49 0 WatchingAgentic AI platform for visual workflows and autonomous agents. Run it with Docker on your...
TOO EARLY 67 14 +14 to +54 -1 Watching🛡️ BLA (bla-agents / Blind LLM Agents) — Open-source Blue Defense MCP server & CLI protect...
REAL 95 92 +4 +49 Quiet climberAn AI live-stream host that runs on one server: voice, face, chat, memory, posts. The base...
TOO EARLY 56 12 +4 to +44 +2 WatchingA free AI assistant and survival library that runs on your own computer, with no internet...
MID 62 51 +11 +6 WatchingLocal multi-AI desktop workspace with official website panels, shared prompts and group co...
MID 42 19 +23 0 WatchingAn easier way to use AI on your own computer. A free, open-source alternative to Claude an...
OVERHYPED 92 61 +30 +1 WatchingIt won't fix your bug. It'll fix you. The debugger that hands the question back instead of...
OVERHYPED 85 32 +53 0 Watchingmulti algo miner for AMD, NVIDIA and Intel gpu's wildrig-multi
MID 17 7 +9 0 SlippingUnlock full speed FP16 tensor core of NVIDIA Turing GPUs
TOO EARLY 51 21 -8 to +33 +1 WatchingExperimental NVIDIA tensor-compute unlock research: guarded TU102 initramfs port, offline...
MID 30 45 -14 +1 WatchingSteering-first neuron explorer for Qwen3.5-4B: push every MLP neuron, keep every rollout,...
TOO EARLY 93 52 +19 to +59 +4 WatchingLow-latency local speech-to-speech voice assistant with Ming-3.0-audio-tiny on AMD and Nvi...
SLEEPER 27 75 -48 0 Watchingllama.cpp on a jailbroken PS5: run any GGUF model natively on the console, starting with Q...
MID 30 14 +15 +4 WatchingYou type. It already knows. Local LLMs that read and think while you type.
SLEEPER 19 85 -66 0 WatchingLearn to sing the songs you love. A free, open-source singing trainer for Mac that listens...
MID 30 30 +0 0 WatchingDeepSeek-V4.1-Flash-1017 EXL3 on 2x DGX Sparks with TensorFold
REAL 59 64 -5 +2 WatchingYour GeForce runs AI at half speed. This fixes it for 20/30/40 series cards. Up to 2× FP16...
OVERHYPED 100 54 +46 +81 SurgingAI jailbreaks: 35 paste-ready AI jailbreak prompts for 25+ frontier and open-weight models...
TOO EARLY 53 8 +4 to +44 +1 WatchingServe large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB c...
REAL 70 74 -4 +6 WatchingBlazing fast Apple Silicon inference engine optimized for Qwen & Nemotron. OpenAI + Anthro...
REAL 88 93 -6 +14 WatchingDeepInfer: High-performance, large-scale inference framework for DeepSeek models
OVERHYPED 86 51 +36 +30 WatchingLocal AI BGM and sound effects for Claude Code and any MCP client. Stable Audio 3 on Apple...
REAL 56 73 -17 +3 Watchingllama.cpp on a jailbroken PS5: run any GGUF model natively on the console, starting with Q...
REAL 56 78 -22 +7 Cooling off1M param OCR model - runs on CPU, 15x faster than tesseract, close to 1B+ param models in...
REAL 76 63 +13 +5 Cooling offDictation that runs on your own computer. Press a key, speak, and the text lands in the ap...
SLEEPER 5 93 -88 0 SlippingXiryn — the open-source personal agent. Does everything Muse, Grok Bot and Dots do, on you...
MID 31 39 -9 0 Cooling offARTEX 한국어판 ·AI 기반 침해사고 대응 기능 추가 (upstream: Autumn-27/ARTEX, AGPL-3.0)
MID 32 51 -19 +2 Watching