Meta enters the crowded AI coding battle with Muse Spark 1.1
Meta enters the crowded AI coding battle with Muse Spark 1.1
Meta 加入拥挤的 AI 编程战场，推出 Muse Spark 1.1
Meta publicly launched a new version of Muse Spark on Thursday, a multimodal AI model designed for agentic coding that aims to compete with similar products offered by OpenAI and Anthropic. Meta 周四正式发布了 Muse Spark 的新版本。这是一款专为智能体（Agentic）编程设计的原生多模态 AI 模型，旨在与 OpenAI 和 Anthropic 提供的同类产品展开竞争。
Spark 1.1, the first version of which was announced in April, can engage in multistep reasoning and handle complex processes, manage digital workflows, and deploy new features in enterprise systems, the company says. 该公司表示，Spark 1.1（其首个版本于今年 4 月发布）能够进行多步推理，处理复杂流程，管理数字工作流，并在企业系统中部署新功能。
Meta is a bit behind its competitors here; Anthropic and OpenAI have offered similar models for quite some time. But that doesn’t mean Meta’s entry into the market isn’t a threat. 在这方面，Meta 比竞争对手稍显落后；Anthropic 和 OpenAI 提供类似模型已经有一段时间了。但这并不意味着 Meta 进入该市场不会构成威胁。
An ongoing source of competitiveness within the AI industry remains the cost of usage, and Meta appears to be offering a competitive rate. Reuters reports that the company will charge $1.25 per million input tokens and $4.25 per million output tokens. That puts it in line with (albeit slightly above) Anthropic’s Claude Haiku 4.5 and OpenAI’s GPT-5.6 Luna. AI 行业持续竞争的一个核心因素依然是使用成本，而 Meta 似乎提供了极具竞争力的价格。据路透社报道，该模型的定价为每百万输入 token 1.25 美元，每百万输出 token 4.25 美元。这一价格与 Anthropic 的 Claude Haiku 4.5 和 OpenAI 的 GPT-5.6 Luna 基本持平（尽管略高一点）。
Meta’s pitch to users is Spark’s ability to handle large agentic workloads, fix bugs, and help with large code migrations — the kind of automation that enterprises are increasingly turning to AI companies to provide. Meta 向用户推销的卖点在于 Spark 处理大规模智能体工作负载、修复漏洞以及协助大型代码迁移的能力——这正是企业日益寻求 AI 公司提供的那种自动化服务。
“Muse Spark 1.1 delivers exceptional performance in personal agentic tasks that require planning and orchestration across a range of external apps and services,” the company wrote in a blog post. 该公司在博客文章中写道：“Muse Spark 1.1 在需要跨多个外部应用程序和服务进行规划与编排的个人智能体任务中，表现出了卓越的性能。”
Meta has released a handful of foundation AI models over the past few years. The Muse Spark release was apparently important enough to compel CEO Mark Zuckerberg to post on X for the first time in three years. Zuckerberg’s last post was in July 2023, around the time the platform rebranded from Twitter to X. 过去几年里，Meta 发布了多款基础 AI 模型。此次 Muse Spark 的发布显然意义重大，以至于促使 CEO 马克·扎克伯格（Mark Zuckerberg）三年来首次在 X（原 Twitter）上发帖。扎克伯格上一次发帖是在 2023 年 7 月，当时该平台正从 Twitter 更名为 X。
In his post, Zuckerberg called Spark “a strong agentic and coding model at a very low price,” noting that the model was “strongest at agentic performance, tool use, and computer use.” Zuckerberg also noted that there was “more to come soon” — implying that the company plans to release additional models. 在帖文中，扎克伯格称 Spark 是“一款价格极低且强大的智能体与编程模型”，并指出该模型“在智能体性能、工具使用和计算机操作方面表现最强”。扎克伯格还提到“更多内容即将到来”，暗示公司计划发布更多模型。
It’s been a big week for AI announcements — particularly for Meta, which also unveiled a new AI image-generation model on Tuesday, dubbed Muse Image. Other releases this week have included a new version of Grok from SpaceXAI and a new family of models from OpenAI, GPT-5.6, that also dropped Thursday. 本周是 AI 领域发布重磅消息的一周，尤其是对 Meta 而言，该公司周二还发布了一款名为 Muse Image 的全新 AI 图像生成模型。本周的其他发布还包括 SpaceXAI 推出的新版 Grok，以及 OpenAI 于周四推出的新系列模型 GPT-5.6。
Suffice it to say that the competition within the AI industry is as healthy as ever, and companies that wish to stand out from their peers have their work cut out for them. 可以说，AI 行业内的竞争依然如火如荼，那些希望在同行中脱颖而出的公司，未来仍面临着艰巨的挑战。