Local AI coding agent you can access from anywhere.
Open-source. Any model. WebRTC remote access. Skills you can share.
Website · Demo Video · Docs · Skills Marketplace · Discord
Linggen is an AI coding agent that runs on your machine — and lets you access it from any device via WebRTC. Start a task on your desktop, check on it from your phone. No cloud hosting, no subscriptions, your models and your data.
curl -fsSL https://linggen.dev/install.sh | bash
lingThat's it. Opens a web UI at localhost:9898.
| Linggen | Claude Code | Cursor | Codex | |
|---|---|---|---|---|
| Runs locally | Yes | Yes | No (cloud) | Cloud-only |
| Any model | Ollama, Claude, GPT, Gemini, DeepSeek, Groq, OpenRouter | Claude only | Multi-model | GPT only |
| Remote access | P2P WebRTC — your machine, any device | Cloud-hosted web app | No | Cloud-only |
| Open source | MIT | No | No | CLI only |
| Skills/extensions | Drop-in SKILL.md files (Agent Skills standard) | Custom slash commands | Plugins | No |
| Web UI | Full web interface with streaming | Terminal + web app | IDE-embedded | Web (cloud) |
| Data stays local | Always — P2P, no relay | Code sent to cloud for web app | Cloud-processed | Cloud-processed |
| Cost | Free + your model costs | $20/mo or API costs | $20/mo | API costs |
Start a coding task on your desktop, monitor it from your phone. No VPN, no port forwarding — peer-to-peer encrypted connection.
ling login # link to linggen.devThen open linggen.dev/app from any browser, anywhere.
Use local models via Ollama, or cloud APIs — Claude, GPT, Gemini, DeepSeek, Groq, OpenRouter. Switch models mid-conversation. Configure fallback chains so work never stops.
Drop a SKILL.md into your project and the agent gains new capabilities instantly. Skills follow the open Agent Skills standard, compatible with Claude Code and Codex.
~/.linggen/skills/my-skill/SKILL.md
Browse and install community skills from the marketplace.
Agents delegate tasks to other agents — each with its own context, tools, and model. Like fork() for AI.
For complex tasks, the agent proposes a plan before acting. Review, edit, or approve — then it executes. Stay in control on high-stakes changes.
Schedule recurring tasks with cron expressions. Code reviews, dependency updates, monitoring — agents self-initiate work on your schedule.
# Install
curl -fsSL https://linggen.dev/install.sh | bash
# First-time setup
ling init
# Start (opens browser)
ling
# Optional: enable remote access
ling loginDrop a markdown file in ~/.linggen/agents/ — available immediately, no restart:
---
name: reviewer
description: Code review specialist.
tools: ["Read", "Glob", "Grep"]
model: claude-sonnet-4-20250514
---
You review code for bugs, style issues, and security vulnerabilities.- Design docs — architecture, specs, and internals
- Full docs — guides and reference
- Skill spec — how to write skills
Contributions welcome. See the design docs for architecture context.
MIT