9 tracked signals on local-models.
Introducing Muse Glimmer
Simon Willison · Aug 10, 2026
Meta releases Muse Glimmer, a 30B open-weights agentic model under Apache 2.0 license
“Muse Glimmer achieves strong success rates on full-task benchmarks including DeepSearch QA, MCP-Atlas, 𝛕-Bench and SWE-Bench, which measure its ability to work within scaffolds, write and debug code, and resolve multi-turn requests from start to finish.”
Memory Harnesses for Long-Running Research Agents — Stefania Druga, Sakana.ai
AI Engineer · Aug 12, 2026
Memory harnesses for local models can solve context rot in long-horizon research agents.
“that makes this issue of dealing with context rot a priority”
The Desktop Frontier — Ahmad Osman, Osmantic
AI Engineer · Jul 21, 2026
Local open-source models will match frontier intelligence on a single consumer GPU within 18 months
“within roughly 18 months we are going to have the equivalent of GLM 5.2 class intelligence running on a single RTX 5090 with 32 GB of VRAM”
Frontier results, on device - RL Nabors, Arize
AI Engineer · Jun 29, 2026
Local on-device models can replace frontier models like GPT-5 and Claude to cut inference costs, latency, and security risks.
“Every time you reach for foundation models like GPT-5 or Claude, it's costing you, your users, and the environment.”
Coding and Personal Agents with Ollama | LIVE145
Microsoft Developer (Build) · Jun 04, 2026
Ollama is launching a privacy-centered cloud service so developers can run open models locally or in the cloud, with agents now its top use case.
“For Ollama, it's the easiest way for developers to get up and running with models, open models specifically, um, on locally and now in the cloud.”
Qwen3.8 27B addition in words
Simon Willison · Oct 04, 2026
Qwen3.8 27B scores only 23.57% on addition-in-words tasks versus GPT-4o's far higher accuracy
“I'm confident GPT-4o didn't cheat and use a calculator, especially since it got so many of the calculations wrong”
Frontier models for the big problems. #aiagents #jetbrains #deeplearning
DeepLearningAI · Aug 24, 2026
Frontier models handle big problems but local models offer control and privacy
“you just want some control, you want some privacy, you just want to keep things local and make my stuff my stuff”
Agent Memory EXPLAINED - Complete Architecture
Hugging Face · Aug 17, 2026
Agent long-term memory requires persistent storage beyond stateless LLM conversational context
“LLMs are stateless machines.”
Take back control of your AI coding workflow
DeepLearningAI · Aug 12, 2026
DeepLearning.AI and JetBrains launch course on flexible, local AI coding workflows
“This course isn't about finding the one right setup. It's about learning how to think about your setup so you can keep making good choices as the tools change.”