质量分 81Bopen-multi-agent.com#business #coding #data #no-code #writing

Open Multi-Agent:TypeScript 多智能体编排框架

Open Multi-Agent 是一个基于 TypeScript 的开源多智能体编排框架,用于构建目标驱动的任务 DAG、并行 LLM 调用和混合模型团队。它提供协调器代理在运行时分解目标、并行执行独立节点并合成类型化结果,支持云、本地或气隙模型。

访问官网

关于这个工具

什么是 Open Multi-Agent?

Open Multi-Agent 是一个基于 TypeScript 的开源多智能体编排框架,专为 AI 工程师和 TypeScript 开发者设计。它允许你构建目标驱动的任务 DAG、并行 LLM 调用和混合模型团队。

核心能力

运行时任务 DAG:协调器代理在运行时将目标分解为任务 DAG,并行运行独立节点,并合成类型化结果。
混合模型团队:在同一团队、同一运行中混合使用不同模型,在需要的地方使用强模型,其他地方使用廉价模型。
本地与私有部署:支持本地 Ollama 端点,可运行整个团队(包括协调器)在小型本地模型上,实现零 API 成本。
Vercel AI SDK 集成:可作为 Vercel AI SDK 应用之上的协调层,保留 AI SDK 用于流式和模型调用。
灵活配置:可为规划器单独配置模型、适配器或额外指令,而不影响工作代理。

适用场景

将目标转化为运行时任务 DAG
自适应客户支持,协调器选择专家并合成证据
为 Vercel AI SDK 应用添加多智能体编排
使用 Ollama 运行本地、私有多智能体团队
在同一个智能体团队中混合使用任意模型

技术细节

采用 MIT 许可证,开源。
提供公共 API。
支持自托管。
学习曲线中等。
商业模式包括免费和免费增值。
运行时依赖三个。
支持云、本地或气隙模型,使用你自己的凭证。

注意事项

一旦授予,bash 将不受沙箱限制,且每个工具结果都会发送到你的模型提供商,因此请有意授予读/执行权限。
定价模式未知。

典型场景

' }, }) Fan-out without dependencies Section titled “Fan-out without dependencies” For MapReduce-style parallelism, use AgentPool
// Blog Which workflows are worth a team of agents
← All use cases// use case Add multi-agent orchestration to a Vercel AI SDK app Keep the AI SDK for streaming and model calls
← All use cases// use case Mix any model in one agent team Use a strong model where it earns its cost and a cheap one everywhere else — in the same team, in one run
← All use cases// use case Run a local, private multi-agent team Keep the data on your machine: open-multi-agent runs against a local Ollama endpoint just like a cloud provider
← All use cases// use case Turn a goal into a task DAG — at runtime In most agent frameworks you draw the task graph by hand
0 latest release MIT license temodar-agent by xeloxa WordPress security analysis platform
02 the approach How open-multi-agent does it

功能与使用场景

' }, }) Fan-out without dependencies Section titled “Fan-out without dependencies” For MapReduce-style parallelism, use AgentPool// Blog Which workflows are worth a team of agents← All use cases// use case Add multi-agent orchestration to a Vercel AI SDK app Keep the AI SDK for streaming and model calls← All use cases// use case Mix any model in one agent team Use a strong model where it earns its cost and a cheap one everywhere else — in the same team, in one run← All use cases// use case Run a local, private multi-agent team Keep the data on your machine: open-multi-agent runs against a local Ollama endpoint just like a cloud provider← All use cases// use case Turn a goal into a task DAG — at runtime In most agent frameworks you draw the task graph by hand0 latest release MIT license temodar-agent by xeloxa WordPress security analysis platform02 the approach How open-multi-agent does it13: Route, Govern, Schedule, and Prove the Run Jul 24, 2026 · 2 min readReleaseOpen Multi-Agent v114: Repair the Plan, Keep the Record Aug 2, 2026 · 4 min readReleaseOpen Multi-Agent v1A coordinator agent decomposes it into a task DAG at runtime, runs the independent nodes in parallel, and synthesizes a typed resultA coordinator plus scoped reviewer agents, with custom tools and token-aware diff compression