VEXI is an open-source AI coding agent built for developers who prefer the terminal. Bring your own API keys, connect OpenAI, Anthropic, and other providers, and accelerate coding, debugging, refactoring, and automation through a lightweight, multilingual, privacy-friendly CLI.
Hey Product Hunt!
I'm excited to share VEXI with the community.
As a developer, I found myself constantly switching between different AI tools, browser tabs, and coding environments. Most solutions were either tied to a specific provider, required subscriptions, or pulled developers away from the terminal where they already spend most of their time.
That's why I built VEXI.
VEXI is an open-source AI coding agent that runs directly in your terminal. It follows a BYOK (Bring Your Own Key) approach, allowing developers to use their own API keys and choose the AI providers they prefer.
Key features:
• Open source and community-driven
• Works directly from the terminal
• Multi-provider support
• BYOK, use your own API keys
• Multilingual support
• Lightweight and easy to set up
• Privacy-friendly workflow
My goal is to give developers a flexible, transparent, and provider-agnostic AI coding assistant without locking them into a closed ecosystem.
I'd love to hear your feedback, suggestions, and feature requests. Every comment helps shape the future of VEXI.
Thanks for checking it out!
GitHub: https://github.com/Elomami1976/vexi
Website: https://vexi.pro
Open-source terminal-first agents are where a lot of real developer workflow will shake out. BYOK + provider flexibility is a good default. What does VEXI do to keep long-running refactors inspectable before changes hit the filesystem?
@sarveshsea Great question, Sarvesh. This was a core design goal.
Vexi keeps long-running work inspectable in three ways:
1. Nothing hits the filesystem without an explicit apply/skip/edit confirmation per change. You see a colored diff before anything is written.
2. Before any edit is applied, Vexi snapshots only the affected files, not a full-tree git mirror, so you can run "vexi undo" or "vexi redo" to step back instantly without touching your real Git history.
3. Every session is recorded and exportable via "vexi replay" as a standalone HTML, so you can review the entire refactor step by step, even after the fact.
The goal is simple: the agent proposes, you stay in control. Would love your thoughts if you try it.
VEXI
memi
Open-source terminal-first agents are where a lot of real developer workflow will shake out. BYOK + provider flexibility is a good default. What does VEXI do to keep long-running refactors inspectable before changes hit the filesystem?
VEXI
@sarveshsea Great question, Sarvesh. This was a core design goal.
Vexi keeps long-running work inspectable in three ways:
1. Nothing hits the filesystem without an explicit apply/skip/edit confirmation per change. You see a colored diff before anything is written.
2. Before any edit is applied, Vexi snapshots only the affected files, not a full-tree git mirror, so you can run "vexi undo" or "vexi redo" to step back instantly without touching your real Git history.
3. Every session is recorded and exportable via "vexi replay" as a standalone HTML, so you can review the entire refactor step by step, even after the fact.
The goal is simple: the agent proposes, you stay in control. Would love your thoughts if you try it.