The latest AI news and updates of Meta — the company behind Llama open-source models, FAIR research, and AI across Facebook, Instagram, and WhatsApp. Covering Meta's latest launches, research, and company news from the past 90 days.
MetaOct 19Meta published the technical report for Muse Glimmer, a 30B open-weight model purpose-built for local agentic tasks. The model runs on a Mac or PC with a single consumer GPU at 4-bit precision. Meta reports that Muse Glimmer leads comparably-sized open models like Gemma4-31B and Qwen3.6-27B on tool use, multi-step planning, and agentic task completion.
MetaOct 19Meta completed the first phase of its gigawatt-scale Prometheus supercluster in New Albany, Ohio, to train the Muse foundational models. The cluster achieved over 85 percent goodput on pretraining and 90 percent on reinforcement learning, with a mean time between interruptions of about 47,000 GPU hours. Meta is scaling the facility toward one million GPUs.
MetaOct 15Meta released Muse Spark Open, an 816B-total parameter model featuring a 1M-token context window and native multimodality across text, image, video, and audio. Built for coding and agentic tasks, the model is optimized to run on two Apple M5 Ultra Mac Studios. It is available via API partners or as downloadable weights under the Muse Model License.
MetaOct 14Meta shared ten research manuscripts produced by medical experts using a human-led research scaffold powered by Muse Spark 1.3 and Muse Wonder. The system accessed five million de-identified electronic health records to accelerate hypothesis generation and manuscript drafting. Researchers reported up to a tenfold productivity gain in best-case scenarios while maintaining full authority over scientific decisions.
MetaOct 12Meta launched the Muse Wonder series, featuring Muse Wonder and Muse Wonder Full, two frontier models designed for personal and professional agentic tasks. Both models introduce Active Memory for context retention and Active Focus for sustained multi-step execution. Muse Wonder is available via the Meta Model API and Muse Code, while Muse Wonder Full is available for Muse Code subscribers.
MetaOct 2Meta shared six mathematics research papers developed by mathematicians collaborating with Muse Spark 1.1 and 1.2 in Thinking Mode. The research, conducted through the meta.ai interface without custom scaffolds, addresses previously open problems in probability, group theory, and physics. Each paper marks AI-drafted passages and includes independent peer review to ensure transparency and scientific rigor.
MetaSep 24Meta unveiled Muse Realtime Avatar, an embodiment technology that turns Muse Realtime Voice into expressive, interactive avatars for live conversations. The system uses a shared speech-token stream and a distilled two-step student model, achieving a 60x reduction in model evaluations. It streams 448x768 video at 25 frames per second, outperforming existing commercial avatar systems in head-to-head preference tests.
MetaSep 8Meta launched Muse, a personal AI agent powered by its Muse Spark 1.3 model, designed to manage tasks within an isolated virtual machine. A host-side Sentinel agent governs all tool actions and network egress to ensure privacy. Meta also opened a bug bounty program offering up to $300,000 for valid security reports, including successful prompt injection attempts.
MetaSep 5Meta’s autonomous research system, AIRA3, placed 8th out of 4,000 teams in a live NVIDIA Kaggle competition, earning a Gold Medal. The system uses an ensemble of models and asynchronous agent coordination to autonomously improve model reasoning. Beyond the competition, AIRA3 has demonstrated capabilities in optimizing production GPU kernels and translating ancient Akkadian clay tablets.
MetaSep 2Meta released Muse Spark 1.3, featuring improved performance on long-horizon agentic and coding tasks. The model now actively collaborates by asking clarifying questions and confirming consequential actions, while reducing hallucination rates. Internal comparisons show a 20% reduction in tool calls and 25% fewer tokens. It is available today in Muse Code and Meta Model API.
MetaSep 1Meta Superintelligence Labs launched Muse Voice Transcribe, a real-time audio perception model for streaming ASR, diarization for over 20 speakers, and endpointing. The model uses reinforcement learning for adaptive delay, balancing speed and accuracy. It ranks first on Artificial Analysis benchmarks for streaming speech-to-text and is available via the Meta Model API, Meta AI for Mac, and Muse Code.
MetaAug 22Meta previewed WildArtifactBench, an internal framework for evaluating AI agents on complex, real-world tasks. Instead of fixed rubrics, it uses win rates and Elo scores from human and agentic preference judges to assess multimodal workflows. Meta released 10 tasks from the benchmark to help measure the practical utility that multimodal agents deliver in diverse deliverable formats.
MetaAug 22Meta released new evaluation results and demonstrations for Muse Spark 1.2, highlighting its visual reasoning and tool-use capabilities. The model now guides robots through unstructured environments, plans subtasks for bimanual manipulation, and generates captions for internal media training data. Muse Spark 1.2 is available via the Meta Model API and Muse Code.
MetaAug 10Meta released Muse Glimmer, a 30-billion-parameter open-weight model optimized for always-on local agent workflows. The model uses 4-bit quantization to fit under 20GB and a DFlash speculative decoding drafter to accelerate token generation on consumer hardware. It is available now under an Apache 2.0 license on Hugging Face, supporting multi-step agentic tasks and multimodal reasoning entirely on-device.
MetaAug 6Meta entered its AI models in five international STEM Olympiads, achieving gold-medal-level performance across all competitions without using external tools. The models secured perfect scores on the theory exams for both the Asian and International Physics Olympiads, alongside gold-medal results in the International Mathematical Olympiad, International Chemistry Olympiad, and the Romanian Masters of Mathematics.
MetaAug 5Meta released Muse Code (beta), a terminal coding agent powered by the new Muse Spark 1.2 model. The agent uses persistent async background agents and an append-only event log to manage long-horizon software engineering tasks with restart-safe execution. Muse Spark 1.2 features co-trained coding optimizations and is available today via the Meta Model API and terminal installation.
Frequently asked questions