Meta Platforms has unveiled Muse Spark 1.1, an upgraded artificial intelligence model designed for coding and advanced agentic workflows, marking the company’s strongest move yet into the increasingly competitive AI developer market.
Released in developer preview on Thursday, Muse Spark 1.1 is positioned as Meta’s most capable model for real-world coding, debugging and complex multi-step tasks. The multimodal model can understand text, images and video while using software and external tools with reduced human intervention, placing it in direct competition with offerings from OpenAI, Anthropic and Google.
The launch also signals a shift in Meta’s AI business strategy. Unlike its previous emphasis on open-weight Llama models, Muse Spark 1.1 is available through the company’s new Meta Model API with usage-based pricing, allowing developers to build commercial applications while giving Meta a new avenue to monetize its AI investments.
Meta first introduced Muse Spark in April through its Superintelligence Labs initiative, which was established to accelerate the company’s frontier AI ambitions. The latest version delivers improved performance for software development, reasoning and agentic tasks while expanding access through a public developer preview.
The release comes amid an intense race among AI developers to deliver increasingly capable coding assistants. Recent launches from OpenAI and other competitors have intensified pressure on major technology companies to demonstrate meaningful advances in reasoning, automation and developer productivity.
Muse Spark 1.1 is also expected to play a broader role across Meta’s ecosystem. The company plans to integrate the model into Meta AI experiences spanning WhatsApp, Facebook, Instagram, Messenger, its AI website and smart glasses, reinforcing its strategy of embedding advanced AI across consumer products.
The announcement follows Meta’s wider push to expand its AI infrastructure, including plans to ramp up computing capacity and deploy custom AI chips as it seeks greater control over the hardware powering its growing portfolio of AI services.
While developers are only beginning to evaluate Muse Spark 1.1, the launch underscores Meta’s determination to compete at the highest level of generative AI. By pairing a powerful coding model with enterprise API access and deep integration across its platforms, the company is positioning itself as a serious challenger in the next phase of AI development.
