Meta 发布 Muse Code，面向大型代码库的终端 AI 智能体

在人工智能应用领域被认为有点落后的Meta，正在努力追赶。 本周，该公司发布了一款新的终端编码代理，旨在为希望在大型软件代码库中处理复杂任务的程序员提供帮助。

目前处于测试阶段的Muse Code能够“完成大型代码库的完整软件工程任务”，Meta首席执行官马克·扎克伯格在周三的一篇社交媒体：https://x.com/finkd/status/2085080750034940201?s=20 帖子中表示。 他补充说，这些任务包括“规划变更、编写代码、验证结果”。

Code可以通过一个命令安装，其背后驱动的是Meta此前发布的编码模型Muse Spark。 它通过启动自己的代理来处理大型项目，这些代理随后会同时工作。

“当任务足够大时，它会展开成在隔离的工作树中并行工作的子代理，”扎克伯格解释道。 “你的工作副本永远不会被触碰。在测试中，我们让它同时为一个游戏构建六个功能，而没有发生冲突。”

此举旨在使Meta在AI实验室同行中更具竞争力，同时成本更低，例如OpenAI及其编码代理Codex，以及Anthropic的Claude Code。

Meta负责人工智能的主管、领导Meta超智能实验室的Alexandr Wang在接受：https://www.wsj.com/tech/ai/meta-releases-coding-agent-to-compete-with-openai-and-anthropic-af87b517?mod=e2twd 《华尔街日报》采访时表示：“我们认为，对于很多工作流程和用例来说，这可能是一个非常不错的选择，尤其是从成本角度来看。”

Meta一直试图通过投入资金：https://www.reuters.com/business/meta-narrows-annual-capex-forecast-ai-buildout-grows-2026-07-29/ 来扩大其人工智能领域的存在。 今年6月，它超越了以使用AI支持广告业务为核心的重点，进入企业AI市场：https://www.reuters.com/business/meta-launches-enterprise-focused-ai-business-agent-automate-daily-operations-2026-06-03/，推出了一款面向客户服务和支持的代理。

加速规模。扩大投资组合。获得实用经验。无论你的目标是什么，Disrupt都可以帮助你。今天最多可节省330美元！

每个工作日和星期天，你都可以获取TechCrunch的最佳报道。

TechCrunch Mobility 是您获取交通新闻和见解的目的地。

初创公司是 TechCrunch 的核心，因此每周获取我们最精彩的报道。

为关键人物提供他们开始一天所需的信息。

Meta, considered a bit of a straggler in the AI harnesses realm, is making strides to catch up. This week, the company released a new terminal coding agent aimed at programmers looking for assistance with complex tasks across large software code bases.

Muse Code, which is currently available in beta, can accomplish “complete software engineering tasks across large repos,” Meta CEO Mark Zuckerberg said in a social media：https://x.com/finkd/status/2085080750034940201?s=20 post on Wednesday. Those tasks include “planning changes, writing code, validating the results,” he added.

Code, which can be installed with a single command, is powered by Meta’s previously released coding model, Muse Spark. It handles large projects by launching its own agents, which then work simultaneously.

“When a job is big enough, it fans out to separate sub-agents working in parallel in isolated worktrees,” Zuckerberg explained. “Your working copy is never touched. In testing we had it build six features for a game simultaneously with no collisions.”

The move attempts to position Meta more competitively, and more affordably, with AI lab peers like OpenAI and its coding agent Codex, and Anthropic with Claude Code.

“We think that for a lot of workflows and a lot of use cases, this can be an incredibly good option, especially from a cost perspective,” Alexandr Wang, Meta’s AI chief who leads Meta Superintelligence Labs, told：https://www.wsj.com/tech/ai/meta-releases-coding-agent-to-compete-with-openai-and-anthropic-af87b517?mod=e2twd the Wall Street Journal.

Meta has been attempting to grow its AI presence by pouring money：https://www.reuters.com/business/meta-narrows-annual-capex-forecast-ai-buildout-grows-2026-07-29/ into development. In June, it expanded beyond its core focus of using AI to support its advertising business and entered the enterprise AI market：https://www.reuters.com/business/meta-launches-enterprise-focused-ai-business-agent-automate-daily-operations-2026-06-03/ with an agent aimed at customer service and support.

Scale faster. Grow your portfolio. Gain practical expertise. No matter your goal, Disrupt can empower you. Save up to $330 toda y!

Every weekday and Sunday, you can get the best of TechCrunch’s coverage.

TechCrunch Mobility is your destination for transportation news and insight.

Startups are the core of TechCrunch, so get our best coverage delivered weekly.

Provides movers and shakers with the info they need to start their day.