跳到主要内容Skip to main content
← Automation

LangflowLangflow

DataStax(原 Langflow AI)DataStax (Langflow AI origin)
开源Open source MITMIT 画布Canvas LLM应用LLM applications DataStaxDataStax 自托管Self-hosted 组件市场Component marketplace partial lifecycle: active 核验Checked 2026-10-30

Langflow 是 使用 MIT 许可的 LLM 画布平台,现归 DataStax—— 产品叙事向企业数据栈(AstraDB 向量线)倾斜。Python 自定义组件是它的扩展主轴。

Langflow is an LLM canvas platform with a whole-repository MIT license, now under DataStax — the product narrative tilts toward the enterprise data stack (AstraDB vector line). Python custom components are its extension axis.

适合与不适合When to use it, when not to

适合:要 MIT 自由度、要 Python 写自定义组件、DataStax/向量库栈用户。 不适合:要开箱企业治理(在 DataStax 托管线);要「LangChain 官方」血统(它是独立近亲)。

A fit for: MIT-level freedom, writing custom components in Python, DataStax/vector-DB stack users. Not a fit for out-of-the-box enterprise governance (on the DataStax hosted line); expecting “LangChain official” lineage (it is an independent cousin).

中文生态China availability 国产 / 国内可直接用 / 备案状态 · 快照domestic / direct access / CAC filing · snapshot 2026-10-21

出品属地Origin
国外出品foreign-made Langflow AI 创立,现归 DataStax(美国)。Founded as Langflow AI, now under DataStax (US).
国内使用Direct access
大陆可直连direct OK docs.langflow.org / langflow.org 直连实测 200(2026-10-21)。docs.langflow.org / langflow.org measured 200 direct (2026-10-21).
备案状态CAC filing
未见于备案名单not in filing list:未见于网信办公示名单(2026-10-21 检索);该产品未面向大陆公众提供服务,备案不适用。not in the CAC public registry (searched 2026-10-21); the product does not serve the mainland public, so filing does not apply.

备案信息以网信办《生成式人工智能服务已备案信息》公示名单为准(按批次滚动更新,本站只抄编号不解读);「大陆可直连」是本站从大陆网络直连实测的结果,不是厂商承诺。GitHub 类站点当日实测有波动,口径写在依据里。Filing status is quoted from the CAC public registry (updated batch by batch; this site copies the number and does not interpret it). "Direct access" is this site's own reachability measurement from a mainland network, not a vendor promise. GitHub-hosted sites fluctuated on the measurement day — the note says so.

固定坐标系Fixed coordinate system

8 个维度dimensions, 与同分区其他对象逐项可比each comparable item by item against others in the same section.

模型与开放条件Model & openness

组件接各家 LLM/embedding/向量库;DataStax 收购后与其向量数据库(AstraDB)线深度整合, 其余 provider 覆盖以官方 docs 为准,未逐条核验。

Components connect to each vendor's LLMs/embeddings/vector stores; after the DataStax acquisition it is deeply integrated with their vector-database line (AstraDB); other provider coverage follows the official docs, not verified item by item.

运行位置Where it runs

开源自托管(pip/uv/Docker)+ DataStax 托管线;GitHub 155,885★、最近推送 2026-10-21(API 核验,非常活跃)。

Open-source self-hosting (pip/uv/Docker) + the DataStax hosted line; GitHub 155,885★, last push 2026-10-21 (API-verified, very active).

本地文件Local files

自托管实例的数据、flow 定义与凭据全部落在自有环境;官方文档给导出/导入 flow 的机制 (细节未逐条核验)。Cloud 托管线数据在 DataStax 云端。

On a self-hosted instance, data, flow definitions and credentials all land in your own environment; official docs describe flow export/import mechanisms (details not verified item by item). The Cloud hosted line keeps data on DataStax's cloud.

关机后的任务Tasks after shutdown

自托管实例常开即承载 API 服务与后台执行;官方给部署文档(细节未逐条核验)。 「电脑关了还能跑」取决于你把实例跑在哪——平台不替你保管关机后的执行。

A continuously running self-hosted instance carries API serving and background execution; official deployment docs exist (details not verified item by item). "Keeps running when your laptop is off" depends on where you run the instance — the platform does not hold executions for you after shutdown.

工具与扩展Tools & extensions

组件市场(marketplace)+ 自定义组件是官方扩展面;MCP 支持状态未逐条核验。

Component marketplace + custom components are the official extension surface; MCP support status not verified item by item.

上下文与记忆Context & memory

flow 内状态与记忆组件是官方上下文面;跨 flow 记忆未逐条核验。

In-flow state and memory components are the official context surface; cross-flow memory not verified item by item.

权限与限制Permissions & limits

开源版基础鉴权;企业治理随 DataStax 托管线——开/商分界未逐条核验。

Open-source edition has basic auth; enterprise governance comes with the DataStax hosted line — the open/commercial boundary not verified item by item.

适合什么任务Best for

适合:要 MIT 许可的 LLM 画布、要 Python 自定义组件、DataStax 栈。 不适合:要开箱企业治理(在托管线)。

Fits: an MIT-licensed LLM canvas, Python custom components, the DataStax stack. Does not fit: out-of-the-box enterprise governance (in the hosted line).

头号误解Top misconception

  • 把它当「LangChain 官方产品」——它是独立项目(现归 DataStax),与 LangChain/LangGraph 是近亲不是同门Treating it as a “LangChain official product” — it is an independent project (now under DataStax), a cousin of LangChain/LangGraph, not a sibling
  • 只看到星数不看归属——DataStax 收购后产品叙事向企业数据栈倾斜(按官方口径建档)Looking only at stars and not ownership — post-acquisition the narrative tilts to the enterprise data stack (filed per official positioning)

未知项清单Known unknowns

  • provider 全清单(未逐条核验)
  • MCP 支持状态(未核验到官方文档)
  • 开源版与 DataStax 托管线的功能分界(未逐条核验)
  • The full provider catalog (not verified item by item)
  • MCP support status (no official documentation found)
  • The feature boundary between the open-source edition and the DataStax hosted line (not verified item by item)

证据来源Evidence sources

判断可回到以下一手源复核。本站核验日Every judgement can be rechecked against the primary sources below. Verified on 2026-10-30,内容更新日content updated on 2026-10-21。

类型Type名称Name链接Link
repolangflow-ai/langflow · 仓库(155,885★,MIT,最近推送 2026-10-21,GitHub API 核验)langflow-ai/langflow · repository (155,885★, MIT, last push 2026-10-21, verified via GitHub API)https://github.com/langflow-ai/langflow
docsLangflow · 官方文档Langflow · official docshttps://docs.langflow.org
docsLangflow · 官方首页Langflow · official homepagehttps://www.langflow.org
changeloglangflow-ai/langflow · Releases(changelog)langflow-ai/langflow · Releases (changelog)https://github.com/langflow-ai/langflow/releases

实测记录Measured runs

本档未跑工作流搭建实测;结论只覆盖平台能力与权限边界。

This site has not completed measured runs. The closest test protocol is the code-side protocols/framework/_protocol.md (the tool-calling and credential-boundary sections are reusable; no workflow-platform-specific protocol exists on this site yet).

相关条目Related entries

  • FlowiseFlowise同为 LLM 画布:MIT 全仓 vs Apache+ee 混合、DataStax 企业线 vs 独立公司——许可与归属是分野。Both LLM canvases: MIT repo-wide vs Apache+ee mixed, DataStax enterprise line vs an independent company — license and ownership are the divides.
  • LangGraphLangGraph另一层 · 框架与运行时Other layer · Frameworks & runtimes

本页由This page is generated from the ai-handbook 数据层生成(CC BY 4.0)。data layer (CC BY 4.0). 方法论与坐标系定义见仓库内For the methodology and coordinate definitions, see METHODOLOGY.md.