Meta发布Muse Spark 1.3多模态模型
Meta新发布的旗舰多模态模型，原生支持计算机使用和多智能体编排，对Agent开发者有直接参考价值。
Muse Spark 1.3
Muse Spark 1.3
Muse Spark 1.3 is trained for agentic workflows and optimized for competitive coding performance. Developers can expect higher first-attempt accuracy and reliable tool calling.
Muse Spark 1.3 专为智能体工作流训练，并针对竞争性编码性能进行了优化。开发者可以期待更高的首次尝试准确率和可靠的工具调用能力。
Start building
开始构建
Documentation
文档
Meet Muse Spark 1.3
认识 Muse Spark 1.3
Trained for long-horizon, agentic workflows
专为长周期、智能体工作流训练
Muse Spark 1.3 tracks context and prior results, works through messy or conflicting inputs, and asks for input when needed.
Muse Spark 1.3 能够追踪上下文和先前结果，处理混乱或冲突的输入，并在需要时请求输入。
Competitive coding performance
竞争性编码性能
Tuned for long-horizon coding workflows, with fewer unnecessary turns and cleaner output. Whether you’re building coding agents or using AI as a development partner, it performs competitively with frontier models across several coding evals.
针对长周期编码工作流进行了调整，减少了不必要的交互轮次并生成更清晰的输出。无论您是在构建编码智能体还是将 AI 作为开发合作伙伴使用，它在多项编码评估中的表现均能与前沿模型相媲美。
Native multimodal perception
原生多模态感知
Muse Spark perceives video, images and documents, and its visual reasoning runs through a real execution environment instead of scripted steps. Feed it a screenshot or a clip and let it build.
Muse Spark 能够感知视频、图像和文档，其视觉推理通过真实的执行环境而非脚本步骤进行。给它一张截图或一段视频片段，让它来构建。
Muse Spark
Muse Spark
Introducing Muse Spark 1.3
介绍 Muse Spark 1.3
Introducing Muse Spark 1.3, with max reasoning for challenging reasoning and agentic tasks and improved real-world usability.
隆重推出 Muse Spark 1.3，具备强大的推理能力以应对具有挑战性的推理和智能体任务，并提升了实际可用性。
Read more
阅读更多
Muse Spark 1.3 benchmarks
Muse Spark 1.3 基准测试
GET STARTED
开始使用
More ways to build.
更多构建方式。
Muse Code Multi-agent coding in your terminal. Built for Muse Spark.
Muse Code 终端内的多智能体编码。专为 Muse Spark 打造。
Start building
开始构建
Meta Model APIDirect, self-serve and affordable access to Muse Spark.
Meta Model API 直接、自助且负担得起的 Muse Spark 访问权限。
Start building
开始构建
OpenRouterSwap Muse Spark into your existing tools and workflows.
OpenRouter 将 Muse Spark 集成到您现有的工具和工作中。
Start building
开始构建
Models and pricing
模型与定价
Model
模型
Context window
上下文窗口
Input (Mtok)
输入 (Mtok)
Cached input (Mtok)
缓存输入 (Mtok)
Output (Mtok)
输出 (Mtok)
muse-spark-1.3-contributorUsed to improve our products.
muse-spark-1.3-contributor用于改进我们的产品。
1M
1M
$0.10
$0.10
$0.002
$0.002
$0.20
$0.20
muse-spark-1.3Not used to improve our products.
muse-spark-1.3不用于改进我们的产品。
1M
1M
$1.25
$1.25
$0.15
$0.15
$4.25
$4.25
Cookbooks and quickstarts
食谱与快速入门
Read documentation
阅读文档
Quick startYour first working request in under five minutes. Point your existing OpenAI SDK compatible client at Meta Model API.
快速入门在五分钟之内完成您的第一个可工作请求。将您现有的兼容 OpenAI SDK 的客户端指向 Meta Model API。
Learn more
了解更多
Multi-agent orchestrationStand up a four profile agentic team that turns a one line product idea into a working SaaS app.
多智能体编排搭建一个由四个配置文件组成的智能体团队，将一行产品创意转化为可用的 SaaS 应用。
Learn more
了解更多
Agent fan-outSpin multiple subagents into isolated worktrees to tackle parallel work without mid-flight collisions.
智能体扇出将多个子智能体旋入隔离的工作树中，以并行处理任务而无需中途碰撞。
Learn more
了解更多
Computer useGive Muse Spark eyes and a pair of hands and it can build software the way you do.
计算机使用赋予 Muse Spark 眼睛和一双双手，它就能像您一样构建软件。
Learn more
了解更多
GitHub agentBuild an autonomous GitHub bot on Meta Model API with Muse Spark and OpenCode.
GitHub 智能体在 Meta Model API 上使用 Muse Spark 和 OpenCode 构建自主 GitHub 机器人。
Learn more
了解更多
Search groundingGive your model live access to the web. Add a single tool to your request to get real time information.
搜索 grounding 赋予模型对网络的实时访问权限。在请求中添加单个工具即可获取实时信息。
Learn more
了解更多
Deploy Muse Spark
部署 Muse Spark
Get started
入门指南
Model documentation
模型文档
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力