Insights, Releases & Local AI Guides
Stay up to date with new Loco releases, local inference benchmarks, architecture deep-dives, and tutorials for running offline AI models.
Z.ai Announces GLM-Legal, an AI Model Built for Legal Work
Z.ai has announced GLM-Legal, a specialised AI model designed for contract review, evidence analysis and regulatory compliance. Built on GLM-5.3-Flash, the model aims to improve how AI handles complex legal workflows.
Meta Muse Glimmer: A New Open AI Model for Local AI Agents
Meta's Muse Glimmer is a 30 billion parameter open weight AI model designed for AI agents, with capabilities spanning tool use, coding, image understanding and multi step workflows. Discover what it means for local AI and the future of private business automation.
JetBrains Releases Mellum 2.1, a New AI Model for Autonomous Coding
Discover JetBrains Mellum 2.1, an open weight AI coding model designed for autonomous coding agents, efficient inference and private local AI deployment.
LightOnOCR-3: How Businesses Can Use Local AI to Process Sensitive Documents Privately
Discover how LightOnOCR-3 enables businesses to extract information from sensitive documents using open weight AI models, with the potential for fully private, on premises processing.
Le Chonk: The Complete Guide to Europe's New 1 Trillion Parameter AI Model
Meet Le Chonk, Mistral AI's enormous new 1 trillion parameter open-weight AI model from France. Learn how Mistral built Le Chonk, what its 1.05 trillion parameters and 49 billion active parameters actually mean, how its 1 million token context window and multimodal capabilities work, why cybersecurity and coding are major strengths, what makes it significant for European AI, and when you will finally be able to download the model yourself.