Home · 3
n8n Labs Releases Nod: An Open-Source Personal AI Agent Running Locally
n8n Labs has launched Nod, an open-source personal AI agent that runs in a local virtual machine to execute app and web tasks on your computer.
Ollama v0.40.16 Adds Ripple Preview 65B, Fixes API Truncation, and Cuts Mac Memory Use
Ollama v0.40.16 adds Ricursive Intelligence's Ripple Preview 65B model, fixes API endpoints, and saves 27 GiB memory for Qwen3.8 Flash Next on Mac.
Dharma-AI Releases 2B Field Extractor Model for Document Parsing
Dharma-AI introduced a 2-billion-parameter model that extracts structured JSON fields from document images, outperforming frontier LLMs in benchmark tests.
Anthropic Releases Claude Code v2.1.318 with Mod Effort Controls and Gateway IAM Attribution
Claude Code v2.1.318 adds effort controls for subagents, AWS IAM role assumption for gateways, and fixes process cleanup and auth lock bugs.
Google Releases Gemini Omni 1.2 Flash with 15-Second Video Generation and 30% Price Cut
Google released Gemini Omni 1.2 Flash with native 15-second video generation, multi-turn editing, and a 30 percent drop in API pricing.
Anthropic Releases Claude Code v2.1.319 with Background MCP Tasks, Gateway Updates, and Diff Fixes
Claude Code v2.1.319 adds terminal MCP task backgrounding, structured gateway logging, automatic permission denials, and diff rendering fixes.
Ollama v0.40.15 Adds Xiaomi MiMo V2.6, AutoTrust Decision Models, and Multi-Request GGUF Serving
Ollama v0.40.15 adds Xiaomi MiMo V2.6, AutoTrust decision models on Apple silicon, and concurrent GGUF request handling.
NVIDIA Releases Four 16B Cosmos 3 Nano Checkpoints for World Simulation and Robotics
NVIDIA open-sources four 16B Cosmos 3 Nano checkpoints for multi-view driving generation, streaming simulation, and robotics action prediction.
Gradium Releases Phonon: Sub-75MB On-Device TTS Models Using Flow Matching
Gradium has released Phonon, a suite of lightweight text-to-speech models running on-device with top ranks on the Open TTS Leaderboard.
Palette: Fine-Tuning a 20B Model for Custom Slide Deck Generation
IBM Research releases Palette, an open-weight 20B model trained on synthetic data and a check-and-repair harness to generate branded slide decks locally.