After more than a year of prioritizing closed models, Meta is reviving its open-weight strategy with the release of Muse Glimmer, a 30-billion-parameter model designed for local, agentic workflows. The move comes as enterprises increasingly seek to run AI closer to their data and maintain tighter control over their infrastructure—a trend driven not only by cost considerations but also by data sovereignty requirements and regulatory compliance pressures that make cloud-only deployments impractical for sensitive workloads.

Released on Monday under the Apache 2.0 license—one of the most permissive open-source licenses, allowing commercial use, modification, and redistribution without copyleft obligations—by Meta Superintelligence Labs, Muse Glimmer is optimized to run on a Mac or PC equipped with a single GPU. The model handles complex, multi-step agentic workloads, including coding, web research, and debugging tasks.

The launch follows just one week after Meta introduced Muse Spark 1.1, a closed model aimed at advanced reasoning and complex agentic tasks. The back-to-back releases signal that Meta intends to stay competitive in the open model market at a time when Chinese AI vendors are gaining traction with price- and performance-competitive alternatives.

Glimmer also marks a partial return to the open model philosophy Meta pioneered in 2023 with the launch of the Llama foundation model, which became one of the most widely downloaded open-weight model families. Since then, however, the company has gradually deemphasized Llama, even after releasing its fourth generation roughly a year ago.

A Hybrid Strategy

Rather than a wholesale reversal, analysts see Meta pursuing a dual-track approach.

“I don’t think Meta is simply switching back to open models,” said Sid Nag, president and chief research officer at Tekonyx. He explained that Meta aims to keep its closed models competitive with offerings from frontier AI labs, while using open models to broaden its reach—particularly among enterprises outside its social media ecosystem. The approach mirrors Google's strategy, where the open Gemma models complement the closed Gemini line.

While the timing of Meta's releases coincides with a surge of competitive Chinese open models from vendors such as Moonshot and Alibaba, Glimmer occupies a distinct niche. At 30 billion parameters, it is far smaller than some popular new Chinese models, which boast a trillion or more parameters.

“This is a very lean, lightweight model that primarily runs on your desktop,” said Arun Chandrasekaran, an analyst at Gartner. “Meta is going smaller, and Meta is going directly toward the edge of the endpoint devices where the AI is running. It’s primarily meant to cover more local desktop-bound agentic workflows, rather than like a more centralized cloud-based model.”

Chandrasekaran noted that Meta has been closely watching how enterprises deploy AI, especially in light of the success of OpenClaw—now under the OpenAI umbrella—and Anthropic's Claude Cowork.

“People want to run these models closest to where the data is, closest to where the workflow is,” Chandrasekaran said. “To that end, Meta wants to be part of that ecosystem, Meta wants to enable that, Meta wants to hopefully empower more workers in the enterprise to run AI more locally.”

Enterprise Demand for Control and Cost Efficiency

Running models locally also reflects a broader enterprise appetite for control. Nag pointed out that many organizations are seeking domain-specific models, which most frontier model providers struggle to deliver—except Anthropic, which has leveraged Cowork to attract enterprises looking for tailored solutions.

Open models also offer compelling inference economics, Nag said. “Training a frontier model costs a lot of money, running millions of enterprise inference workloads costs even more over time,” he explained. “By releasing weights, Meta lets customers bear much of the infrastructure cost instead of operating every workload themselves.”

Despite these advantages, Chandrasekaran cautioned that Meta still has significant work ahead to win meaningful enterprise adoption. “It needs to build that platform layer on top of the model because that’s what enterprises want,” he said. “Meta should also provide better capabilities around security and legal indemnification for enterprises.”

As the competitive landscape intensifies—with both Chinese open models and U.S. frontier labs vying for enterprise attention—Meta's dual strategy of lean edge-deployable open models alongside heavyweight closed systems positions it across multiple segments of the AI market simultaneously.