Meta launched Muse Spark 1.1, a multimodal AI model for enterprise coding tasks, on Thursday with pricing ($1.25 per million input tokens and $4.25 per million output tokens) competitive with rivals like Anthropic and OpenAI. Although Meta is late to this market, the model is positioned around agentic performance and tool use for handling complex workflows. CEO Mark Zuckerberg signaled more models are coming soon.
Summaries like this, in your inbox every morning.
Sign up free →What happened
Meta publicly released Muse Spark 1.1 on Thursday, a multimodal AI model for agentic coding (AI that plans and executes tasks autonomously) that can handle multistep reasoning, manage digital workflows, and deploy features in enterprise systems. Pricing is $1.25 per million input tokens and $4.25 per million output tokens.
Why it matters
Meta is entering a crowded market where Anthropic and OpenAI already offer similar tools, but the company is competing on cost — its pricing is in line with (albeit slightly above) Anthropic's Claude Haiku 4.5 and OpenAI's GPT-5.6 Luna. Enterprises increasingly seek AI automation for bug fixes and large code migrations, so Spark's pitch for agentic performance and tool use may appeal to business users looking for cost-effective solutions.
What to watch
CEO Mark Zuckerberg posted about the launch on X for the first time in three years, calling Spark a "strong agentic and coding model at a very low price." He also signaled that "more to come soon," hinting at additional model releases. This week alone has seen new versions of Grok from xAI and OpenAI's GPT-5.6 family, underscoring intense industry competition.
Meta's entry into the AI coding-agent market this week reflects the intensity of competition among major AI vendors. The company was apparently far enough along with Spark that CEO Mark Zuckerberg felt compelled to announce it publicly—his first post on X in three years, since July 2023. This suggests the launch carries strategic weight for Meta's AI positioning.
The timing is notable: within days, both xAI and OpenAI released new model versions (a new Grok and the GPT-5.6 family), indicating the industry is in a phase of rapid capability announcements. Meta's pricing strategy—undercutting or matching competitors on per-token cost—points to a deliberate choice to compete on affordability rather than performance differentiation alone. For enterprises deciding between coding-agent tools, cost parity combined with Meta's emphasis on agentic performance and tool orchestration may influence procurement decisions.
Zuckerberg's hint that "more is to come soon" suggests Meta is not treating Spark as a one-off release but as part of a broader model expansion strategy. The company has already released Muse Image (an image generation model) earlier in the week, indicating a multi-capability push across foundational AI models.
No comments yet. Be the first to share your thoughts!
Log in to join the discussion
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack
