updated 62 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++ HTML JavaScript Jupyter Notebook Python Swift TypeScript
Topic: All agent agent-skills ai audio claude-code comfyui cuda cursor deep-learning elevenlabs ffmpeg gemma image-generation jev large-language-models llm llm-agents macos open-source python pytorch qwen self-hosted speech-to-text stable-diffusion text-to-image text-to-speech transformer tts video-generation
Distill a good Claude Code / Codex session into a skill an open-weight model can follow to...
MID 42 29 +13 - WatchingKobiMusic'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 WatchingTraining Reasoning Models in Simulation to Natively Use Real Robots
MID 42 53 -11 +12 WatchingPro Tools 2026 desktop setup with multitrack editing, MIDI sequencing, VST hosting and a p...
MID 42 47 -5 0 WatchingA 4.5M-parameter two-voice English text-to-speech model, trained from scratch
SLEEPER 42 83 -41 +33 WatchingMiniMax-H3-Image, an open-weight model for 16-megapixel image generation and editing.
TOO EARLY 63 27 +0 to +42 +6 WatchingTraining code for PAN2, a open foundation world model for physical intelligence
TOO EARLY 57 9 +7 to +47 +1 WatchingDownload and explore time-series datasets through one standardized format.
REAL 91 80 +11 +1 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 WatchingPublic code release for the paper "The Surprising Effectiveness of Deleting Weights in LLM...
TOO EARLY 56 18 -1 to +41 +2 WatchingTernary 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 WatchingHumanoid whole-body motion tracking with one differentiable world model.
TOO EARLY 74 43 +3 to +45 +14 Quiet climberNaive-N0.5-Flash is an open-weight 309B MoE model with 15.5B active parameters, built for...
TOO EARLY 62 0 +23 to +57 +13 WatchingOpenAmika: MekaHime's AI companion Amika, open source. Live2D avatar, real-time voice, lon...
REAL 59 74 -15 0 Watching:sunrise: A comprehensive, cutting-edge open-source framework for next-generation Transfor...
REAL 62 77 -15 +2 WatchingA frontier model edits another model's weights by hand (Gemma 4 12B, EN->IT tense). Plus a...
MID 22 25 -3 0 WatchingCode for Latent Sparse Attention: Towards Subquadratic Ultra-Long Context Attention
SLEEPER 30 68 -38 0 Watching💃 Free CC0 mannequin pose reference library for AI image tools (Nano Banana, ChatGPT, Comf...
MID 59 41 +18 +4 WatchingAn NES CPU compiled into the weights of a stock Qwen3.5 model
TOO EARLY 87 8 0 WatchingThe first open-source model for code-driven video creation.
TOO EARLY 65 14 +12 to +52 +8 WatchingOne-click character swap for videos: an open-source, local Higgsfield Genjutsu alternative...
TOO EARLY 70 32 +4 to +46 +21 Watching🛡️ BLA (bla-agents / Blind LLM Agents) — Open-source Blue Defense MCP server & CLI protect...
REAL 95 92 +4 +49 Quiet climberBuild AI text detectors whose decisions are mechanistically grounded and human-auditable (...
TOO EARLY 90 8 +1 WatchingAn easier way to use AI on your own computer. A free, open-source alternative to Claude an...
OVERHYPED 92 61 +30 +1 WatchingWeVisOne-VL is a powerful VLM by the WeChat Vision Team at Tencent, unifying object-level...
TOO EARLY 61 12 +10 to +50 +7 WatchingGVS5H: Five Qwen3.8-27B Models Match Claude Fable 5 on LiveCodeBench Hard | Fable 5 Level...
TOO EARLY 87 13 +1 WatchingBuild AI phone agents on your own computer and test every change with an AI caller. Open s...
SLEEPER 17 65 -48 0 SlippingIndependent reproduction of SuperMimic (arXiv:2610.28480) on mjlab/MuJoCo Warp, written fr...
MID 29 34 -5 0 WatchingTurkish voice design with VoxCPM2 on a 4 GB GPU: describe a voice, pick a variant, read te...
REAL 62 69 -7 +9 WatchingMing-3.0-tts is an open-source TTS model family with 0.5B dense and 7.9B-A1.3B MoE variant...
TOO EARLY 57 23 -3 to +39 +4 Watchingvibe coded supermimic within 1h, LAFAN passed within 1.5 gpuh training
MID 64 50 +14 0 WatchingAI jailbreaks: 35 paste-ready AI jailbreak prompts for 25+ frontier and open-weight models...
TOO EARLY 53 8 +4 to +44 +1 WatchingBattlezone Online: a free 3D multiplayer tank game in the browser (desktop, phone or VR) t...
SLEEPER 1 80 -79 0 WatchingMake any NPC in any game talk back with Fish Audio: skills, a self-hosted voice server and...
MID 13 38 -25 0 WatchingOfficial repository for MOSS-Image, a 5B-parameter base DiT model with high-quality text r...
TOO EARLY 79 27 +16 to +58 +11 Cooling offClaudemy: an open-source alternative to Udemy and Coursera. Give Claude a topic, it resear...
MID 59 39 +19 0 Cooling offCorpus legislativo brasileiro em larga escala e modelo Transformer causal treinado do zero...
MID 16 5 +11 +2 Watching1M param OCR model - runs on CPU, 15x faster than tesseract, close to 1B+ param models in...
REAL 76 63 +13 +5 Cooling offOfficial implementation of Video Prediction Policy 2: Predict Better, Act Better
MID 25 24 +1 +1 WatchingEdgeRock: Strong Reasoning at Constant Memory. Multimodal models with O(N) generation and...
TOO EARLY 72 15 +18 to +58 +48 WatchingTyped decisions from a frozen language model and a small physics engine: calibrated probab...
SLEEPER 18 87 -70 +3 Watching一句话让 AI 帮你做产品宣传片!把产品名或网址告诉 Cursor、Claude Code、Codex 等 AI 编程工具,它会自己调研产品、画出多种风格样图让你挑选、截取 4K...
LEANS MID 31 1 -9 to +26 0 WatchingRSI-Anything: Unlocking Recursive Self-Improvement with World Models
TOO EARLY 64 7 +16 to +56 +2 WatchingScam With Your Friends Web: unofficial open-source browser remake. Co-op AI call center ga...
MID 16 38 -21 0 WatchingTiny, fast and accurate Turkish text-to-speech. 8.6M parameters, 0.92% WER on Freya-TR-Eva...
MID 9 27 -17 0 WatchingOfficial implementation of LittleBit (NeurIPS 2025) and its follow-up LittleBit-2 (ICML 20...
TOO EARLY 75 4 +28 to +67 +107 AcceleratingQwen's most powerful open-source image generation model
TOO EARLY 58 1 +17 to +52 +1 Watchingverl/HybridFlow: A Flexible and Efficient RL Post-Training Framework
MID 13 6 +7 +1 WatchingQwen and DeepSeek low-VRAM LoRA training with GGUF base model, compatible with Transformer...
SLEEPER 4 48 -44 0 WatchingContrastive Language Models (CLMs) are a new class of System One model trained with a cont...
MID 26 11 +15 -1 WatchingPre-registered independent eval of TypeSafe Jev against a nano-class LLM, a frontier LLM,...
TOO EARLY 42 11 -9 to +31 0 WatchingReal-time browser demo for Talk Like You
MID 3 5 -2 0 WatchingKandinsky 6.0 Video: Foundation Models for Synchronized Video and Audio Generation
SLEEPER 2 34 -31 0 SlippingNon-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over...
REAL 81 94 -14 +54 WatchingOne AI trade decision every Monad block. Jev on Kuru MON-USDC.
MID 26 5 +22 +2 WatchingDetect youtube sponsor segment with live audio and transcript powered by Jev
MID 2 3 -1 0 WatchingAn intuitive console agent to fine-tune AI models with little to no experience required!
MID 5 4 +1 0 WatchingGLM-5.3-Flash EXL3 on 2x DGX Spark with TensorFold
SLEEPER 24 68 -44 -1 WatchingEvolving Multimodal Inference Programs: LLM-guided evolutionary search for accurate and ef...
MID 2 9 -7 0 WatchingODS V3 Pre-Release: Public testing and refinement ahead of the official V3 launch. Turn yo...
REAL 62 78 -16 +24 Watching🤗 Transformers: the model-definition framework for state-of-the-art machine learning model...
REAL 82 98 -17 +7 Quiet climberOpen-source framework for the research and development of foundation models.
SLEEPER 55 91 -35 +2 WatchingFrontier CoreML audio models in your apps — text-to-speech, speech-to-text, voice activity...
SLEEPER 55 87 -32 +1 WatchingKronos: A Foundation Model for the Language of Financial Markets
TOO EARLY 85 2 +24 Gone quietTraining-free language model modification: Remove censorship, suppress slop, reduce sycoph...
TOO EARLY 98 67 +24 to +46 +120 SurgingUncensored Open-source alternative to AI video platforms — Free AI image & video generatio...
REAL 88 69 +19 +90 Quiet climberLocal UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek...
REAL 81 98 -17 +64 Quiet climberWorld's first open-source, agentic video production system. 12 production pipelines, 100+...
REAL 90 85 +5 +205 Surging利用 AI 大模型和自动化工作流,根据主题或关键词一键生成高清短视频。Generate HD short videos from a topic or keyword with a...
REAL 90 84 +6 +129 SurgingVoiceStudio is the open-source, fully-local ElevenLabs alternative — voice cloning, voice...
REAL 99 99 +0 +483 Surging