Meta enters the crowded AI coding battle with Muse Spark 1.1
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Meta launched Muse Spark 1.1 publicly on Thursday, a multimodal AI model for agentic coding aimed at competing with Anthropic and OpenAI's similar offerings.
- Spark 1.1 is priced at $1.25 per million input tokens and $4.25 per million output tokens (per Reuters), putting it slightly above Anthropic's Claude Haiku 4.5 and OpenAI's GPT-5.6 Luna.
- The model handles multi-step reasoning, complex workflows, bug fixes, and large code migrations; Meta calls its performance on personal agentic tasks "exceptional."
- Mark Zuckerberg posted about Spark on X for the first time in three years, calling it "a strong agentic and coding model at a very low price" and signaling "more to come soon."
- The launch caps a packed AI release week: Meta also unveiled Muse Image on Tuesday, SpaceXAI shipped a new Grok, and OpenAI dropped GPT-5.6 the same day.
Why it matters: Meta is entering a market where Anthropic and OpenAI already have established coding products, but its near-bottom pricing challenges the cost calculus for enterprises comparing agentic coding vendors. Zuckerberg's rare personal promotion—his first X post in three years—signals Meta's strategic bet that price-plus-capability can carve share against entrenched rivals.
