Week of Sep 14 to Sep 20, 2026
Tools of the Week
Every Monday this page reshuffles and puts a different tool forward from each of the 27 categories in the directory. It is a way to surface the good work that never reaches the top of a ratings list. 27 picks are live right now, and the next rotation lands Monday, September 21.
CUDA-accelerated Gaussian splatting rasterization library with Python bindings from the Nerfstudio team.
Platform for building stateful AI agents with advanced long-term memory and self-improvement.
Free markerless motion capture system that works with ordinary cameras and no special hardware.
Self-hosted alternative to GitHub Copilot using open-source models.
Kubernetes-native platform for deploying ML models to production.
Minimalist deep learning framework in under 10,000 lines of code.
All-in-one video model unifying reference-to-video, video-to-video, and masked editing.
Evaluation platform that benchmarks LLMs on over 70 datasets with distributed execution.
Whisper with word-level timestamps and speaker diarization
Simple and effective multi-object tracking using every detection box.
Framework and CLI for building, converting, and analyzing computer vision datasets across many formats.
Tool for fine-tuning LLMs with various configurations
Multi-platform package manager for Stable Diffusion UIs.
Latest Stable Diffusion model with improved text rendering and composition.
Baidu's Apache-2.0 family of ten language and vision-language models with a heterogeneous MoE design.
Run large language models locally with a simple CLI interface
PyTorch-native platform for large-scale generative AI pretraining with FSDP2, 4D parallelism, and float8.
Open-source music generation model for creating full songs with vocals and accompaniment.
Simple framework for state-of-the-art NLP by Zalando Research.
Multilingual document OCR toolkit with line detection and layout analysis.
High-performance vector database for similarity search
Rev's ASR and diarization models trained on 200,000 hours of human-transcribed English audio.
Real-time voice cloning and TTS model with 1.2B parameters by MetaVoice.
SvelteKit chat frontend for OpenAI-compatible APIs that powers Hugging Face's HuggingChat.
Fast Rust search engine API with typo-tolerant full-text and hybrid semantic vector search.
Twin-backbone model generating short video clips with synchronized speech and sound in one pass.
Community fork of RVC with additional features and optimizations.