Hi Devs!
We’re a small group of open-source AI enthusiasts working within a big company, trying to build something meaningful. Today, we’re super excited to share Nexent with you — a project that makes it incredibly easy to go from a natural language prompt to a working AI agent.
No drag-and-drop. No scripts. No code. Just a prompt → plan → execute.
🧠 What Is Nexent?
Nexent is a zero-code, open-source platform that automatically generates intelligent agents from natural language. These agents aren’t just fancy chatbots — they can:
- Understand your goals
- Plan multi-step tasks
- Use external tools and APIs
- Manage memory and context
- Return results in structured formats
Nexent agents combine models, tools, and data into fully functional intelligent workflows — all triggered by a simple prompt.
🛠️ Try It in 4 Lines
Yes, it’s that simple to get started. Just clone, install, and run:
git clone https://github.com/ModelEngine-Group/nexent.git
cd nexent/docker
cp .env.example .env # fill only nessasary configs
bash deploy.sh
And you’ll get access to:
- A web interface for running agents
- Built-in tools and examples
- Support for local model plugins
- MCP tools integration
✨ Why We Built Nexent
As developers working inside a large company, we saw a lot of exciting LLM and AI tooling — but most of it required complicated orchestration, scripts, or internal glue code.
We wanted to build something different:
- No-code but not limited
- Extensible via simple plugins (MCP-based)
- Fast to prototype, but strong enough for production
- Open-source and self-hosted, always
🔍 Example Use Case: Knowledge Assistant AI Agent (Multimodal + Source-Aware)
Nexent powers a knowledge assistant agent that goes far beyond basic Q&A. It can:
- 🔎 Search internal docs and the web
- 📚 Trace source citations for every response
- 🧠 Summarize or synthesize across multiple formats
- 🖼️ Present information in text, images, and charts
- 🔧 Use both LLMs and custom tools to answer complex domain-specific queries
Whether it's for research, compliance, or decision support — this agent acts as a smart, explainable layer on top of your knowledge base and APIs.
👀 Want to See It in Action?
the GitHub repo:
👉 https://github.com/ModelEngine-Group/nexent
We’re actively looking for feedback, contributors, and early adopters to help shape the direction of this project.
🙌 How You Can Help
If you find this project interesting, here’s how you can support us:
- ⭐ Star the repo and give us a look!
- 📣 Share it with your network or community
- 🧩 Build a plugin or a new agent use case
- 🐛 File an issue, bug, or feature request
- 🎯 Suggest an integration or application idea
We’re just getting started, and we’d love to build this with you.
🎁 Oh, and a Few Extras...
We offer:
- 🏅 Open-source contributor badges
- 🎁 Swag/gifts for active contributors
- 🧠 Public acknowledgments on our site and GitHub
- 📦 Ready-to-use agent templates and demo configs
Thanks for reading — and we hope you give Nexent a try!
Feel free to comment here or reach us via GitHub or our contact in the README.
The future of intelligent automation starts with a prompt.
Let’s build it — together.
Top comments (0)