Meta has been moving fast in August 2026, and it is not slowing down. On August 5, the company quietly launched Muse Spark 1.2 alongside a brand new product called Muse Code. This is a serious signal that Meta is going all-in on competing with OpenAI's Codex and Google's coding tools in the developer AI space.
Muse Code Is Meta's Answer to AI-Powered Development
Muse Code is a terminal-based coding agent powered entirely by Muse Spark 1.2. It is built to write code, run it, verify the results, and fix errors autonomously in a loop without the developer having to step in at every stage.
Meta is pricing it aggressively at just $1.25 per million input tokens, which is significantly cheaper than comparable tools from OpenAI and Anthropic. The model is currently in beta, but developers who have tested it are reporting strong performance on multi-step coding tasks, particularly in Python and JavaScript.
Reuters confirmed the launch on August 5, and the reaction from the developer community on GitHub and Reddit has been largely positive. Muse Spark 1.2 itself is described as a multimodal reasoning model with major improvements in tool use, computer interaction, and agentic chaining of tasks.
Meta AI Is Now a Full Task Automation Platform
Separate from Muse Code, Meta's main AI assistant has also seen a major capability expansion. Powered by Muse Spark 1.1, Meta AI can now connect to your email and calendar, build task plans, create presentation slides, and handle recurring workflows autonomously.
Meta announced in late July that it is expanding the beta of a business-focused AI assistant inside Meta Ads Manager to all advertisers globally. This means any business running ads on Facebook or Instagram now has access to an AI co-pilot that can analyze campaign performance, suggest budget adjustments, and generate ad creatives automatically.
For a walkthrough of what Muse Spark 1.2 and Muse Code actually do in practice, this video is worth watching:
Meta Launches Muse Code and Muse Spark 1.2
Meta is clearly trying to build an AI ecosystem that stretches from consumer chat all the way to developer tooling and business automation. The strategy is working.
Meta's Q1 2026 revenue hit $56 billion, with AI tools driving a significant portion of that growth. The company that was written off as a laggard in AI just 18 months ago is now one of the most active players in the space.
Quick Links:
