
Comfy-Org Makes Boogu-Image's 10B Open-Source Image Model Plug-and-Play
Comfy-Org repackages the new Boogu-Image-0.1 model family into drop-in ComfyUI-ready files, making the 10B Apache-2.0 image model instantly accessible to node-graph workflows
20 stories

Comfy-Org repackages the new Boogu-Image-0.1 model family into drop-in ComfyUI-ready files, making the 10B Apache-2.0 image model instantly accessible to node-graph workflows

Resume-Matcher is a free, self-hostable AI tool that rewrites your resume to match any job description and beat ATS ranking algorithms

Google DeepMind and the UK government launch an AI planning prototype to cut householder application decisions from 8 weeks to 4, targeting 1.5 million new homes by 2029.

NVIDIA's cuTile Python brings tile-based GPU kernel programming to Python, eliminating the need for C++ and manual thread management in CUDA 13.1

NVIDIA's open-source kvpress library unifies 20+ training-free KV cache compression methods into a single pip install, slashing long-context LLM memory costs by up to 50%.

Bindu wraps any AI agent in one function call, giving it cryptographic identity, A2A protocol compliance, and built-in crypto payments out of the box.

NVIDIA's Cosmos 3 fuses vision reasoning, video generation, and robot action prediction into one open omnimodal world model for physical AI.

Salvatore Sanfilippo (antirez), creator of Redis, ships a purpose-built C inference engine that runs a 284B frontier model locally on a 128GB MacBook with a 1M-token context window.
Y Combinator's Garry Tan open-sources GBrain, a Postgres-backed AI agent memory system with a self-wiring knowledge graph, 43 skills, and overnight 'dream cycles' that compound knowledge while you sleep.
OpenClaude brings Claude Code's agentic workflow to 200+ models — OpenAI, Gemini, DeepSeek, Ollama, and more — all from one open-source CLI

Transformer Engine v2.15 ships FlashAttention 4, MXFP8 attention, and deeper NVFP4 support for Blackwell GPUs — pushing 4-bit pretraining closer to production

CC Switch is a free, open-source desktop app that lets you manage and hot-swap AI providers across Claude Code, Codex, Gemini CLI, and more from a single GUI

xAI's Grok Imagine Video 1.5 exits preview with native audio, faster generation, and the top spot on the image-to-video leaderboard

Paper2Any is an open-source Python toolkit that turns research PDFs, screenshots, or plain text into editable PPTX slides, scientific figures, draw.io diagrams, and academic posters in one click.

Cursor announces Origin, an agent-native Git hosting platform built to handle the throughput demands of parallel AI agents, taking direct aim at GitHub

Krea released a 12B diffusion transformer trained from scratch with open weights, focusing on aesthetic diversity and reference-driven style control over raw benchmark scores.
Lucebox hand-tunes CUDA kernels for consumer GPUs, hitting 207 tok/s on Qwen3.5-27B on a single RTX 3090 — no H100 required

A long-standing precision bug silently corrupted model buffers like RoPE tables during mixed-precision training — DeepSpeed 0.19.2 finally fixes it
Midjourney is unveiling its first-ever physical product at a live San Francisco event — and the only clue is that you can go inside it.
Google's official Agent Skills repo ships 13 Google Cloud skills for Claude Code, Cursor, and Antigravity — solving context bloat with on-demand expertise