Models & Labs
Meta's Muse Glimmer Model Now Available
Ollama Bloghigh confidence
Why it matters
- →Muse Glimmer enhances local AI agent capabilities with a large context length and multimodal support.
- →The model's optimization for Apple Silicon via Ollama's MLX engine offers significant performance improvements.
- →Its open-source release under Apache 2.0 increases accessibility for developers working on advanced AI applications.
Meta Superintelligence Labs has launched Muse Glimmer, a 30 billion parameter multimodal model, available on Ollama. Designed for agent workloads, it supports a 128K+ context length and runs efficiently on Apple Silicon using Ollama's MLX engine. The model is released under the Apache 2.0 license and can be used for coding agents and personal assistants. This release enhances AI accessibility for developers, particularly on Apple hardware, with improved performance features like DFlash and image input support.
Read originalThe story around this
TopicMeta Releases Glimmer AI ModelCooling
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1Meta Releases Muse Spark Model
Lev Selector · May 15, 2026 · Related
- 2Gemma 4 Boosts Speed with Multi-Token Prediction
Ollama Blog · June 29, 2026 · Related
- 3Meta's Muse Image Model Debuts Strongly
The Rundown AI · July 8, 2026 · Related
- 4Meta Unveils Muse Image AI Model
The AI Daily Brief · July 9, 2026 · Related
- 5Meta Launches Muse Spark 1.1 for AI Coding
TechCrunch AI · July 9, 2026 · Related
- 6Meta Launches Muse Spark 1.1
AI Explained · July 10, 2026 · Related
- 7Meta AI Launches Muse Spark Agent
Matt Wolfe · July 31, 2026 · Related
- 8Meta Releases Muse Spark 1.2 with New Features
The AI Daily Brief · August 8, 2026 · Related
What happened next
- 9Meta Introduces Muse Spark 1.3
Matt Wolfe · September 4, 2026 · Related
- 10Meta Launches Muse, a Personal AI Agent
WIRED AI · September 8, 2026 · Related
- 11Meta's Muse AI assistant launches on Mac
TechCrunch AI · September 18, 2026 · Related
- 12Meta unveils standalone Muse Charm AI hardware
The Verge AI · September 24, 2026 · Related
- 13Meta Muse agent security failures and enterprise push
The Verge AI · September 30, 2026 · Same story
Follow this story
Open the full story →Meta unveils Muse Glimmer: a 30B parameter multimodal model
10 developments
- Aug 10 · Hugging Face Blog
Meta unveils Muse Glimmer: a 30B parameter multimodal model - Aug 10 · Ollama Blog
Meta's Muse Glimmer Model Now Available (This article)↳ Muse Glimmer is now available on Ollama via MLX for Apple Silicon and licensed under Apache 2.0. - Aug 10 · AI News
Meta Muse Glimmer enables local AI on consumer GPUs - Aug 10 · Sam Witteveen
Meta releases Muse Glimmer 30B model↳ Muse Glimmer is an open agentic model available on Hugging Face. - Aug 10 · TechCrunch AI
Meta Releases Open-Weight Glimmer AI Model↳ Glimmer uses an Apache 2.0 license - Aug 11 · The Rundown AI
Meta returns to open-source with Muse Glimmer - Aug 11 · The Rundown AI
Meta Embraces Open Source with Muse Glimmer - Aug 12 · Fireship
Meta Releases Muse Glimmer Model - Aug 14 · TechCrunch AI
Meta Releases Open-Weight AI Model Glimmer - Aug 14 · TechCrunch AI
Meta Releases Glimmer AI Model for Public Use
More in Models & Labs
© The Verge AIModels & Labsother
Google launches Guided Vision in Gemini Live
Google is bringing real-time audio scene description to Android via Gemini Live, directly challenging Apple’s VoiceOver Live Recognition. This feature targets users with low vision by providing immediate audio cues and follow-up Q&A capabilities for physical objects. It integrates deeply into the accessibility ecosystem through TalkBack, moving beyond simple text reading to contextual environmental awareness. The move signals a shift toward multimodal AI as a standard utility for daily navigation rather than just a novelty.
The Verge AI
© TechCrunch AIModels & Labsagents
AWS releases open-source Strands Decider 2B
Amazon’s Strands Decider 2B joins the growing wave of decision models designed to replace heavy LLMs for simple routing tasks. Built on Qwen3.5-2B, it outputs calibrated choices with confidence scores rather than generating text, offering a cheaper, faster alternative for agentic workflows. The release signals AWS’s push into specialized agent infrastructure, aiming to solve the latency and cost bottlenecks of general-purpose models. While TypeSafe’s Jev pioneered this space, Amazon’s entry brings enterprise-grade credibility and open-source accessibility to a niche that is rapidly filling with experimental clones.
TechCrunch AI
© Sam WitteveenModels & Labsmodels
Google Announces Gemini 4 Argon with 1M Token Output
Google is pushing the boundaries of context windows with Gemini 4 Argon, a new model capable of generating up to one million tokens in a single response. This isn't just about reading long documents; it's designed for complex agentic workflows where the AI must produce extensive codebases or detailed reports without truncation. Early benchmarks suggest it aims to reclaim top-tier intelligence status against competitors like GPT-6, specifically targeting tasks that require sustained reasoning and massive output generation. The shift from 64K caps to a million-token horizon fundamentally changes how developers might architect multi-step autonomous systems.
Sam Witteveen