Weave不需编码的 AI 工作流建立工具，使企业能够通过整合多个大型语言模型 (LLM) 和连接提示建立流程...

4.8 (5)

审阅者 Daniel Nikulshyn·更新 2026年6月

No-Code Workflow Automation LLM Integration Business Automation AI Orchestration Prompt Chaining

概览

W&B Weave 是一个可观察性和评估平台，帮助跟踪和改进大型语言模型（LLM）应用程序。Weave 提供了跟踪、收集指标以及使用 LLM 评估器和自定义评分器评估应用程序响应等工具。主要功能包括跟踪会话、LLM 调用和工具调用，以及自定义代理的手动检测。该平台支持与热门的SDK和框架集成，以及自定义代理可观测性。Weave提供了Python和TypeScript库，用于安装和使用该平台。它托管在Weights & Biases（W&B）上，需要一个W&B账户和API密钥进行身份验证。用户可以在 Weave UI 中跟踪对大型语言模型的调用，查看输入和输出，并查看代理指标。Weave 可以实现 LLM 应用的自动化和评估，但它并不是一个如其名称所暗示的无代码 AI 工作流构建工具。

主要功能

代理跟踪和指标收集
自定义代理观察性
LLM 跟踪和评估
OpenTelemetry span 支持
Weights&Biases (W&B) 集成
Python 和 TypeScript 库

价格

模型: Free
分类: 标做管球解析
评分: 4.8 / 5 (5)

使用场景

自动化多步业务操作

建立不需编码的工作流程，链式连接多个 LLM，以自动化可重复的业务流程而无需编写代码。

在一条管道中集成多个 LLM

连接和调度不同的大型语言模型以在单个工作流程中结合其独特优势，以执行特定任务。

简化提示链式

可视化链接提示，以便输出从一个步骤传递到下一个步骤，使团队有能力通过复杂的 AI 驱动序列。

优点 & 缺点

优点

支持流行的 SDK 和 harness
提供自定义代理观察性
与Weights&Biases (W&B) 集成用于数据收集
跟踪和指标收集功能
使用 LLM 判定和自定义评分器评估应用响应

缺点

不具有不需编码的 AI 工作流建立特征
生成助手响应可能包含错误
需要 W&B 账户和 API 密钥进行身份验证

对决战绩

在万神殿中参与了 1 对决。

第1

第2

第3

Last battle

#4
Observability Showdown — July 25, 2026
Jul 25, 2026 · #4 of 9

评测

4.8

5 个评分的平均值。

登录以留下评测。

Wei Chen

Mar 23, 2026

Use it every day

Honestly didn't expect to like it this much. The dashboard is exactly what I needed, and the value for money is strong. but I reach for it almost every day now and it just clicks.

George Papadakis

Feb 7, 2026

Compared a few options

Evaluated this against two competitors. Where it wins: the core workflow and it is genuinely easy to set up. Where it lags: the docs could be deeper. On balance the feature set — especially the dashboard — justifies the 4 stars for our use case.

Ahmed Saleh

Dec 10, 2025

Compared a few options

Evaluated this against two competitors. Where it wins: the onboarding and the value for money is strong. Where it lags: pricing gets steep at scale. On balance the feature set — especially the onboarding — justifies the 5 stars for our use case.

Marcus Bell

Oct 16, 2025

Compared a few options

Evaluated this against two competitors. Where it wins: the integrations and it saves real time. On balance the feature set — especially the dashboard — justifies the 5 stars for our use case.

Fatima Zahra

Sep 27, 2025

Compared a few options

Evaluated this against two competitors. Where it wins: the integrations and the value for money is strong. On balance the feature set — especially the API — justifies the 5 stars for our use case.

问答

What observability features does Weave provide for custom agents?

Weave supports manual instrumentation of custom agents, capturing OpenTelemetry spans, tracing session and tool calls, and displaying metrics in the Weave UI. You can also apply LLM judges or custom scorers to evaluate agent responses.

Asked by Tomáš Novák · Jun 9, 2026

What are the prerequisites for using Weave?

You need a Weights & Biases (W&B) account and an API key for authentication. After installing the Python or TypeScript library, you can start tracing LLM calls and viewing data in the hosted Weave dashboard.

Asked by Rasheed Osman · Jun 10, 2026

How does Weave integrate with existing LLM SDKs and harnesses?

Weave offers Python and TypeScript libraries that hook into popular LLM SDKs and harnesses, automatically tracing calls, inputs, and outputs. The integrations let you collect metrics and evaluate responses without writing custom instrumentation.

Asked by Dumisani Ndlovu · Apr 17, 2026