News
Meta Enters the AI Coding Assistant Race with Muse Spark 1.1
Meta has released Muse Spark 1.1 to developers, a multimodal model built for agentic coding with a million-token context window, priced below OpenAI's offerings and close to Anthropic's. Mark Zuckerberg broke a three-year silence on X to announce the launch.
Contents
Meta opened public access to Muse Spark 1.1 on Thursday, a model built specifically for agentic coding. It marks the company's first serious move into the commercial AI coding assistant market, where Anthropic and OpenAI have been competing head to head for months.
Muse Spark 1.1 is a multimodal model designed to handle multi-step reasoning, debugging, large codebase migrations, and computer-use interfaces on desktop, mobile, and in the browser. Meta says the model can maintain context through long-running tasks and run subtasks in parallel via subagents.
Priced Below the Competition
At $1.25 per million input tokens and $4.25 per million output tokens, Muse Spark 1.1 lands close to the pricing of Anthropic's Claude Haiku 4.5 and OpenAI's GPT-5.6 Luna, though still above Anthropic's cheapest offerings. It's a clear signal that Meta wants to compete primarily on the cost of running large, repetitive agentic tasks in enterprises, rather than on raw single-response quality.
For developers, that means a third viable price point alongside Anthropic and OpenAI when building agentic tools, which matters in practical terms, not just marketing terms, as token bills keep climbing for engineering teams.
Zuckerberg Breaks His Silence
The launch mattered enough to Meta that Mark Zuckerberg posted on X for the first time in three years, describing Muse Spark as a strong agentic and coding model at a very low price. He also signaled more model releases are coming soon.
a strong agentic and coding model at a very low price - Mark Zuckerberg, CEO of Meta
A Shift Away from Open Models
Meta has built its reputation on open models in the Llama family, available for free download and modification. Muse Spark 1.1, by contrast, reaches developers only through a commercial, pay-per-token API, the same model OpenAI and Anthropic have long used. That signals the company now treats coding tools as a business that needs to generate direct revenue, not just a way to build an ecosystem around free models.
The AI coding assistant market is already crowded, and Meta is arriving late compared with Anthropic's Claude Code or tools built on GPT-5.6. The company is looking to build its edge on price and on handling large, corporate-scale automation workloads, rather than on quality advantages in individual tasks.
What It Means for Polish Teams
For Polish software companies and dev shops using agentic tools, the arrival of a third major price player could put pressure on existing vendors to cut their rates. It's also worth watching whether Muse Spark makes its way into integrations popular in Poland, like Cursor or IDE plugins, since that will really determine its adoption beyond Meta's own API.
Sources: Meta enters the crowded AI coding battle with Muse Spark 1.1 (techcrunch.com), Meta jumps into AI coding market in effort to chase Anthropic and OpenAI (cnbc.com), Meta prices Muse Spark 1.1 API at $1.25/$4.25 per million tokens (aiweekly.co)

