-2.png?w=512)
Meta is taking its AI rivalry into one of the industry's most competitive new markets: autonomous software development.
The company has launched Muse Code, its first dedicated coding agent, powered by the new Muse Spark 1.2 model. The tool is designed to inspect codebases and tackle broader software-engineering tasks as Meta moves into a market already crowded with offerings from OpenAI, Anthropic, and Google.
Business Insider reported that Muse Code will cost $1.25 per million input tokens, $0.15 per million cached input tokens, and $4.25 per million output tokens.
The launch puts Meta alongside rivals that have spent months building AI software engineers rather than conversational assistants. It also signals a broader shift in enterprise AI, where success is increasingly measured by how much real work an AI can complete — not simply how well it answers questions.
A brief look inside Meta’s new AI tool
Muse Code is designed to handle agentic software engineering, with Meta positioning it as a tool to tackle more “complete software engineering tasks,” according to CNBC.
The launch also marks the debut of Meta Superintelligence Labs' first coding product, the AI division led by chief AI officer Alexandr Wang. Via CNBC's coverage, Wang said the coding tool is available with a pay-as-you-go option.
Wang also said the pay-as-you-go tier is cheaper than mainstream coding AI platforms while hinting at strong adoption of Muse Code without giving any figures. The only drawback for the contributor tier is that users would have to opt in for their code to be used for model training.
That signals Meta is trying to close the gap with more established coding-agent platforms.
Meta is expanding its AI revenue playbook
Muse Code also appears to be the newest member in Meta's effort to turn its multibillion-dollar AI push into revenue.
While the company has long emphasized how AI improves recommendations across Facebook and Instagram, it has also recently highlighted productivity gains within the business. In July, during Meta’s last earnings call, Meta’s chief, Mark Zuckerberg, told investors that AI is helping engineers ship products faster and automate parts of the software development process.
By entering the market for AI coding agents — which appears to be an increasingly competitive enterprise AI segment — Meta is targeting businesses willing to pay for tools that can reduce software development costs and improve engineering productivity.
AI coding is becoming AI's newest performance benchmark
Meta's arrival should give developers another serious option in an already crowded coding-agent market, and greater competition could put downward pressure on prices while pushing vendors to improve capabilities faster.
But Muse Code's bigger test will be adoption. Meta has spent heavily to catch the leaders in frontier AI; coding agents now give it a market where success can be measured not only in benchmark scores, but in whether developers are willing to pay for the work its models can actually complete.
Related News: Meta is expanding how business-shared customer data can personalize Facebook content and Meta AI responses, raising new privacy and governance questions for users and enterprises.