Meta enters the crowded AI coding battle with Muse Spark 1.1
Meta unveiled Muse Spark 1.1, a multimodal AI model designed to automate complex coding workflows and enterprise system operations. The system handles multistep reasoning and can orchestrate tasks across multiple applications, positioning itself as a direct competitor to offerings from OpenAI and Anthropic despite Meta entering the space somewhat later than these incumbents.
The model's pricing structure—$1.25 per million input tokens and $4.25 per million output tokens—aligns competitively with existing alternatives. Meta's chief executive marked the occasion with a social media post, his first in three years, emphasising the model's capabilities in agentic task handling and external tool integration whilst signalling additional releases to come. The announcement exemplifies the intensifying competition among major AI firms to capture enterprise automation market share.
- Meta launches Muse Spark 1.1, a multimodal AI model for automated enterprise coding tasks including bug fixes and large-scale code restructuring
- Pricing at $1.25/$4.25 per million input/output tokens matches competitors; CEO's rare social post emphasises tool use and agentic capabilities
- Release signals escalating competition in AI, with OpenAI, Anthropic, and now Meta all offering similar enterprise automation models