Open Source - Page 2 of 17
Open-source AI projects, community innovations, collaborative development, and freely accessible AI tools and frameworks.
325 articles
View complete article list
1,478 AI deliberations show persona, not model, shapes epistemic behavior
Here's the thing: the Consilium Protocol treats disagreement between AI models as a clue, not a bug.
Memory OS adds weekly decay scanner and 0.92 similarity dedup to prune stale data
Why does this matter? Because the default memory in Hermes Agent—already capable of remembering across sessions—might not cut it for heavier workloads.
Mistral AI rebrands LeChat to Vibe, positioning it as a full AI work agent
Mistral AI has taken its chatbot “Le Chat” and given it a new name—Vibe—while recasting it as a full‑blown work assistant.
Stability AI releases Stable Audio 3 with diffusion and higher‑noise training
Why does this matter? Because Stability AI just opened the doors to its newest audio‑generation suite, Stable Audio 3, and the weights are now publicly available.
D&B rebuilds 642 million‑business database after AI agents hit limits
Why did D&B have to start from scratch? The answer lies in a data architecture that was never meant for autonomous agents.
Experimental MLX Delegate Enables PyTorch Models on Apple Silicon GPUs
Apple Silicon has become a popular platform for running large language models locally. Until now, ExecuTorch users on macOS were stuck with CPU‑only backends like XNNPACK or the AOTI Metal backend. The new ExecuTorch MLX delegate changes that.
Cohere launches Command A+ 218B sparse MoE, runs on two H100 GPUs using QAD
Cohere just put Command A+ on the table, an open‑source model aimed squarely at enterprise‑grade agentic workflows.
PyTorch Docathon 2026 merges 150+ PRs, adds API and ExecuTorch docs
The PyTorch Docathon 2026 ran from May 5 through May 19, drawing more than 260 registrants and about 30 active contributors.
Cohere open-sources Command A+ with 128k token window, 48‑language support
Cohere is making its most capable language model publicly available. The Canadian AI firm released Command A+ under the Apache 2.0 license, turning a 218‑billion‑parameter mixture‑of‑experts system into open‑source code you can download from Hugging...
Google to roll out AI agents this week, adopting OpenClaw features
Google is finally putting AI assistants into everyday use. At I/O 2026 the company unveiled a suite of agents that can pull data, plan events, and skim your inbox and calendar—all while running in the background.
Google AI agents expand on Alerts, moving beyond simple notifications
At Google I/O 2026 the company unveiled a new layer of “agentic” functionality built into Search. Instead of a one‑off answer, users can now spin up, tweak and run multiple AI agents that keep tabs on topics they care about around the clock.
Google launches three AI subscription tiers at USD 10, adds Gemini Omni and 3.5 Flash
Google reshaped its AI subscription lineup at I/O 2026, swapping daily prompt caps for a consumption‑based billing approach.
Four AI supply-chain attacks in 50 days expose release pipeline red‑team gaps
Four AI supply‑chain attacks in just 50 days have shone a light on a blind spot that most vendors overlook: release‑pipeline red‑team coverage.
SDOF Adds Two Defensive Layers via Intent Router and StateAwareDisp
Multi‑agent orchestration tools like LangChain, LangGraph and CrewAI excel at routing tasks through graph‑based pipelines, yet they rarely enforce the stage constraints that real‑world business processes demand.
Oppo open-sources X‑OmniClaw agent that scrolls, captures, reads prices in‑app
Oppo’s Multi‑X team has just put an Android AI agent on GitHub. The project, called X‑OmniClaw, lets a phone’s camera, screen and voice work together to perform tasks inside real apps—without sending any data to a remote server.
GraphBit Employs Typed Functions and Rust Engine for Reproducible Orchestration
Why does this matter? Traditional agentic LLM setups let the model decide each step, which often leads to stray routing, endless loops or results you can’t reproduce.
Osaurus, an open‑source Mac tool, delivers personal AI with local and cloud models
Why does this matter? Because a new open‑source tool is giving Mac users a way to run AI without handing over every prompt to a cloud service.
2026 Data Science Jobs Favor Professionals Managing AI Agents and Business Skills
The data‑science field moves at breakneck speed. In 2026, newcomers often feel like they’re drinking from a firehose—juggling Python, cloud platforms, and the latest machine‑learning models all at once.
NVIDIA and SAP introduce NemoClaw blueprint to add trust to specialized agents
SAP customers who need custom AI agents now have a clearer route to production. NVIDIA’s NemoClaw—a reference blueprint for building and deploying autonomous agents—has been dropped into Joule Studio, giving development teams a structured path from...
AntAngelMed: 103B-parameter open-source medical model on 1/32 MoE
A Chinese research team just put AntAngelMed on GitHub. It’s an open‑source language model aimed squarely at medical tasks, and the developers claim it’s the biggest and most capable of its kind so far.
📚 Featured Resources & Reviews
🎓
No Code MBA Course Review
Learn to build AI-powered apps without coding. Our comprehensive review of No Code MBA's course.
🔧
AI Tools & Resources
Curated collection of AI tools, courses, and frameworks to accelerate your AI journey.
📬
Weekly AI Digest
Get the week's most important AI news delivered to your inbox every week.