Repolode
Submit
LazyAGI

LazyAGI/LazyLLM

A low-code Python framework for building multi-agent LLM applications, covering RAG, chatbots, fine-tuning and one-click deployment.

View on GitHubWebsite
Stars
3,891
Forks
430
Open issues
77
Last commit
yesterday

What is it?

LazyLLM is a low-code development tool for building multi-agent large language model applications. It provides prebuilt flows such as pipeline, parallel, diverter, and loop, so developers can assemble workflows from functions and modules without writing much orchestration code.

The framework targets a prototype-to-production cycle: build a working app quickly, analyze failing cases with task-specific data, then iterate on algorithms or fine-tune models. It offers a unified interface for online model services from several providers and for locally deployed models served through inference frameworks such as vLLM and LightLLM, and includes RAG components like Document, Retriever and Reranker.

Deployment is part of the design. Applications can be started with a single command, a lightweight gateway handles wiring between submodule services, and the app can be packaged as an image for Kubernetes. The same code can run on bare metal, development machines, Slurm clusters or public clouds.

Key features

  • 01Low-code workflow assembly from functions and modules using pipeline, parallel, and loop flows.
  • 02Unified interface for online model APIs and locally deployed models.
  • 03Built-in RAG components including document parsing, retrieval, and reranking.
  • 04One-click deployment with a lightweight gateway and image packaging for Kubernetes.
  • 05Fine-tuning support with automatic selection of frameworks and model splitting strategies.
  • 06Cross-platform execution across bare metal, Slurm, and public clouds.

Use cases

  • Building a retrieval-augmented question-answering bot over your own documents.
  • Prototyping a multi-agent chatbot that routes requests to chat, speech, or image models.
  • Turning a prototype pipeline into a deployable service on Kubernetes.
  • Comparing online and local models behind the same application code.

Quick start

  1. 1

    Install lazyllm with pip

    pip3 install lazyllm
  2. 2

    Install all dependencies (optional)

    lazyllm install full
  3. 3

    Set your OpenAI API key as an environment variable, then start a chatbot with the bundled command

    lazyllm run chatbot

Things to know

  • Local model inference needs at least one inference framework such as lightllm or vllm, which typically requires a GPU.
  • Windows and macOS installs require a separate setup tutorial from the docs.

In collections

Multi-agent orchestration frameworksOpen-source AI agent frameworks & apps

Topics

#ai-agent#llm#rag#multi-agent#fine-tuning#low-code#python#inference#agents#data

FAQ

What is LazyLLM?
A low-code Python framework for building multi-agent LLM applications, covering RAG, chatbots, fine-tuning and one-click deployment. LazyLLM is a low-code development tool for building multi-agent large language model applications. It provides prebuilt flows such as pipeline, parallel, diverter, and loop, so developers can assemble workflows from functions and modules without writing much orchestration code.
Is LazyLLM free to use? What license does it have?
LazyLLM is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is LazyLLM still maintained?
Actively maintained. The last commit to LazyLLM was on Nov 6, 2026.
What can I use LazyLLM for?
Building a retrieval-augmented question-answering bot over your own documents. Prototyping a multi-agent chatbot that routes requests to chat, speech, or image models. Turning a prototype pipeline into a deployable service on Kubernetes. Comparing online and local models behind the same application code.
How do I install LazyLLM?
Install lazyllm with pip: pip3 install lazyllm Install all dependencies (optional): lazyllm install full Set your OpenAI API key as an environment variable, then start a chatbot with the bundled command: lazyllm run chatbot
Where can I get LazyLLM?
The source code and installation instructions are on GitHub: https://github.com/LazyAGI/LazyLLM

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects