Meta is significantly boosting its ambitions in the field of artificial intelligence. The company has introduced Muse Code, its first AI assistant designed for programming on macOS and Linux computers. The new product is aimed directly at tools like OpenAI Codex, Claude Code or GitHub Copilot, and according to Mark Zuckerberg, it can not only write code, but also independently solve large development tasks.
You might be interested in
Muse Code runs directly in Terminal and does not require a separate application. After a simple installation, it can analyze large projects, suggest changes, edit source code, verify results, and use multiple specialized AI agents simultaneously. If its work is interrupted, thanks to a continuous recording of all operations, it will pick up exactly where it left off.
But the most interesting thing is the pricing policy. The standard tariff costs $1,25 per million input tokens and $4,25 per million output tokens. However, Meta offers a significant discount to those who agree that their input and generated code can be used to further train their AI models. In the so-called Contributor mode, the price drops to $0,10 per million input tokens and $0,20 per million output tokens, or about a twentieth of the regular price.
You might be interested in
It is this model that is causing the most controversy. While some developers welcome the significantly lower costs, others point out that sharing their own source code with Meta may not be acceptable, especially when working on commercial or sensitive projects. The introduction of Muse Code also confirms that the battle for AI assistants for programming is gaining momentum. While OpenAI, Anthropic or GitHub have been offering similar tools for a long time, Meta has been absent in this segment until now. With Muse Code, this is changing and the competition in the market of AI tools for developers is a bit fiercer.
