Meta Launches Muse Code AI Programming Agent to Challenge OpenAI, Anthropic
Translated from Spanish, summarized and contextualized by DistantNews.
At a glance
- Meta has launched Muse Code, its first AI agent for software programming, aiming to compete with OpenAI and Anthropic.
- The tool, integrated with Meta's Muse Spark 1.2 model, is available in beta for macOS and Linux systems.
- Muse Code operates on a pay-per-use model, with API pricing set at $1.25 per million input tokens and $4.25 per million output tokens.
Meta has entered the competitive AI programming assistant market with the launch of Muse Code, its first artificial intelligence agent specifically designed for software development. This move positions the tech giant to challenge rivals like OpenAI and Anthropic, which already offer similar tools.
Muse Code is integrated with Meta's Muse Spark 1.2 AI model and is currently available in a beta phase. Developers can install and run this coding agent on macOS and Linux operating systems. The tool operates as a terminal-based assistant, allowing programmers to interact with AI for coding tasks directly within their development environment.
The company has structured access to Muse Code through a commercial pay-per-use scheme. According to Alexandr Wang, Meta's Director of AI, the pricing for the application programming interface (API) remains consistent with the Muse Spark 1.1 version. Costs are calculated based on the volume of information processed. Specifically, Meta has set a rate of $1.25 for every million input tokens directed to the model and $4.25 for every million output tokens generated by the tool.
The introduction of Muse Code follows Meta's recent technological releases, including Muse Spark 1.1, designed for deep reasoning capabilities, and Muse Image, a graphic generator. These advancements indicate Meta's ongoing commitment to expanding its AI offerings across various domains.
the prices of the application programming interface (API) are maintained at levels equivalent to those of the Muse Spark 1.1 version, calculating costs according to the volume of information processing.
Originally published by El Nacional in Spanish. Translated, summarized, and contextualized by our editorial team with added local perspective. Read our editorial standards.