Understand the project
Practical overview
When it fits
Use Lagent when you need Python-level control over agent construction, message flow or conversation state, rather than a ready-to-use hosted chat service.
Who it is for
- Python developers building custom LLM-based agents.
- Teams prototyping multi-agent applications or custom message flows.
Documented uses
- Connect an agent to a language model and send it structured messages.
- Retain, inspect and reset the messages stored in an agent's memory.
- Customize how conversation history and few-shot examples are assembled for a model.
Before you choose
Requirements & limits
The project is available as a public GitHub repository and lists an Apache-2.0 license. The README documents installation from source with pip.
- The documented workflow is Python code and editable installation from the repository.
- The example configures a VllmModel and a Qwen model path; model setup is a separate consideration from installing Lagent.
- The README includes examples for inspecting and resetting message memory and customizing aggregation.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Source-based practical guide
Install Lagent from its source repository
Clone the repository and install Lagent in editable mode.
Before you start
- A working Git and Python environment with pip.
Steps
Clone https://github.com/InternLM/lagent.git.
Change into the lagent directory.
Run pip install -e .
Expected outcome
The repository README's source-installation procedure is complete.
Guide limitations
- This setup procedure does not configure or run a model-backed agent.
Evidence and transparency
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 11 October 2026
- Original website title
- InternLM/lagent · Official homepage title · retained in its source language