Repolode
Submit
ModelEngine-Group

ModelEngine-Group/nexent

A zero-code platform that generates production-style AI agents from natural-language descriptions, with built-in tools, memory and orchestration.

View on GitHubWebsite
Stars
5,995
Forks
744
Open issues
284
Last commit
yesterday

What is it?

Nexent is a platform for building AI agents without writing code or drawing workflows by hand. Users describe the agent they want in plain language, and the platform assembles it from unified tools, skills, memory and orchestration components, with constraints and feedback loops built in.

The project supports any OpenAI-compatible model provider and covers LLM, embedding, vision-language, speech-to-text and text-to-speech models. It includes a personal-grade knowledge base for 20+ document formats with retrieval and citations, MCP tool integration, agent-to-agent (A2A) collaboration, layered user and user-agent memory, and agent version rollback.

Nexent can be deployed with Docker Compose for individuals and small teams, or on Kubernetes with Helm for enterprise environments. Offline image packages are provided for air-gapped installs. Multi-tenancy and role-based access control are included.

Key features

  • 01Generates agents from natural-language requirements without orchestration or drag-and-drop.
  • 02Works with any OpenAI-compatible provider across LLM, embedding, VLM, STT and TTS models.
  • 03Personal knowledge base supporting 20+ document formats with citations.
  • 04Supports MCP tool extensions and agent-to-agent (A2A) collaboration.
  • 05Layered memory and agent version management with rollback.
  • 06Multi-tenancy and role-based access control.

Use cases

  • Building internal assistants that answer questions over private documents.
  • Prototyping task-specific agents by describing them in plain language.
  • Deploying an agent platform on private infrastructure with Docker or Kubernetes.
  • Coordinating several agents that call each other through the A2A protocol.

Quick start

  1. 1

    Clone the repository

    git clone https://github.com/ModelEngine-Group/nexent.git
  2. 2

    Enter the project directory

    cd nexent
  3. 3

    Deploy with Docker (prerequisites: Docker 24+ and Docker Compose v2+)

    bash deploy.sh docker

Things to know

  • Requires substantial resources: at least 4 CPU cores, 8 GiB memory and 40 GiB disk for Docker deployment.
  • Kubernetes deployment needs at least 16 GiB memory and Helm 3+, which suits enterprise setups rather than small installs.
  • Self-hosting involves a multi-service stack with configuration and infrastructure work.

Topics

#agent#llm#rag#mcp#multi-agent#no-code#knowledge-base#kubernetes#agentic-ai#agentic-framework

FAQ

What is nexent?
A zero-code platform that generates production-style AI agents from natural-language descriptions, with built-in tools, memory and orchestration. Nexent is a platform for building AI agents without writing code or drawing workflows by hand. Users describe the agent they want in plain language, and the platform assembles it from unified tools, skills, memory and orchestration components, with constraints and feedback loops built in.
Is nexent free to use? What license does it have?
nexent is open source under the MIT license. Check the license terms before using it in commercial products.
Is nexent still maintained?
Actively maintained. The last commit to nexent was on Oct 20, 2026.
What can I use nexent for?
Building internal assistants that answer questions over private documents. Prototyping task-specific agents by describing them in plain language. Deploying an agent platform on private infrastructure with Docker or Kubernetes. Coordinating several agents that call each other through the A2A protocol.
How do I install nexent?
Clone the repository: git clone https://github.com/ModelEngine-Group/nexent.git Enter the project directory: cd nexent Deploy with Docker (prerequisites: Docker 24+ and Docker Compose v2+): bash deploy.sh docker
Where can I get nexent?
The source code and installation instructions are on GitHub: https://github.com/ModelEngine-Group/nexent

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

Related projects