Back
Features
Everything NeurAI does
The full list, by category. The edge isn't any single feature — it's the integrated combination, with the agent understanding everything from local code to production logs.
An agent that runs
It doesn't suggest code — it acts. Reads the project context and works end to end.
- Reads, writes, and edits project files
- Runs terminal commands and scripts
- Full git operations (branch, commit, diff, merge)
- Background execution with real-time log tracking
- Live task plan — a checklist that updates as the agent progresses
- Subagents: delegates parts of a task to specialized agents
- Optional prompt refiner for more precise instructions
- Cancellation and fine-grained control of running processes
Your servers, over SSH
The agent works on your real infrastructure, not just a sandbox.
- Secure SSH connection to remote servers
- Installs dependencies, starts services, and restarts processes
- Reads and investigates production logs
- Diagnoses incidents right on the affected machine
- Streams output from long-running commands
- Credentials stored securely (encrypted vault)
Project-tied deploys
From edit to production without switching tools.
- Deploy target configured per project (separate from the dev environment)
- Deploys triggered from the chat
- History of deploy runs
- Real-time tracking of the result
24/7 AI monitoring
It tells you when something goes down — and investigates the cause.
- HTTP endpoint and log monitors
- Automatic likely-cause analysis on anomalies
- Notifications when something goes off-script
- Scheduled tasks for recurring routines
Interactive preview with visual annotation
Runs your app inside the platform and fixes what you point at.
- Frontend preview running inside NeurAI
- Visual inspector: click an element and describe the fix
- The agent applies the change to your project's real code
- Smart proxy for apps with varied routes and assets
Local models, zero cost
Your data on your machine, no paid API key.
- Connects Ollama and LM Studio through the Bridge
- Messages with local models are UNLIMITED on every plan
- No data leaves your machine when using a local model
- Multi-LLM: switch between local models and APIs without changing your flow
- No vendor lock-in
All from your phone
Manage and debug from anywhere.
- Responsive interface, built mobile-first
- Follow sessions and processes from your phone
- Act on incidents without opening your laptop
Account, teams & security
Product infrastructure ready for serious use.
- Authentication with 2FA
- Encrypted secrets vault
- Plans and limits managed by admin
- Ticket-based support and a usage-vs-plan page
- Recurring security audits
Start free
Unlimited local models on every plan · no card to start