Meta has officially launched Muse Spark 1.1 on Thursday, marking its entry into the competitive AI coding landscape. The multimodal AI model is designed for agentic coding, aiming to provide enterprises with enhanced capabilities for managing digital workflows and deploying new features. This launch comes as Meta seeks to catch up with established players like OpenAI and Anthropic, who have already introduced similar models.
Muse Spark 1.1 Features and Capabilities
The Muse Spark 1.1 model is touted for its ability to engage in multistep reasoning, manage complex processes, and assist with large code migrations. According to reports, the model can effectively handle agentic workloads, which are increasingly demanded in today’s AI-driven enterprise environments. Meta positions this model as a cost-effective solution, charging $1.25 per million input tokens and $4.25 per million output tokens.
- Input token cost: $1.25/million
- Output token cost: $4.25/million
- Agentic capabilities: bug fixing, code migration
Competitive Landscape in AI Coding
Despite being late to the market, Meta's Muse Spark 1.1 poses a potential threat to its competitors. The pricing is competitive, though slightly higher than Anthropic’s Claude Haiku 4.5 and OpenAI’s GPT-5.6 Luna. The AI landscape remains vibrant, with various companies unveiling new models this week, including Meta’s own Muse Image model and Grok from SpaceXAI.
Mark Zuckerberg's Enthusiasm for Muse Spark
The significance of Muse Spark 1.1 was underscored by Mark Zuckerberg, who made his first post on X in three years to promote the model. In his post, he described Spark as “a strong agentic and coding model at a very low price,” emphasizing its capabilities in agentic performance and tool use. Additionally, Zuckerberg hinted at more developments to come from Meta in the AI space.
🤖 This article was rewritten by Feed and Figures' editorial AI from a report originally published by TechCrunch. Facts and quotes are preserved from the original; the rewrite focuses on clarity and structure. For the unedited original, see the source link below.
