Skip to content

Your AI agents, deployed.

One Discord message. Zero YAML. Your agent goes live in 10 seconds.

No Dockerfiles. No CI pipelines. No infrastructure babysitting.

Free forever · No credit card · No config files

$ @nott deploy github.com/you/my-agent
Reading repo · Detecting framework · Building
✅ live! https://my-agent.menot.app · 8.3s

How it works

One mention. Full outcome.

No YAML. No config files. No dashboards to babysit.

01
@nott deploy auth-agent to staging

@mention in Discord

Type what you need in any Discord forum post. Deploy, debug, review — in plain English. No commands to memorize.

02
Cloning repository
Building container
Deploying to staging

We build and ship

Repo connected. Container built. Environment configured. Agent running — zero YAML, zero config, zero waiting.

03
Deployed 4.2s
https://staging.auth-agent.dev
Build: 3.8s Image: 142MB

Result in Discord

Real-time progress right in the thread. Logs, metrics, and a green checkmark when the job is done.

Features

Everything you need. Nothing you don't.

Built for engineers who ship, not engineers who configure.

Zero-Config Deploy

Push to any repo, @mention the agent, watch it ship. No Dockerfile required. No ENV management.

Discord Native

Everything lives in Discord threads. No new dashboards, no new tools. Your team already knows where to look.

Auto-Scaling

Agents scale with load automatically. Idle workspaces hibernate in minutes. You pay only for what runs.

Real-Time Logs

Streaming logs in the Discord thread as the agent works. No SSH, no kubectl logs, no log aggregation setup.

Team Collab

Shared workspaces, shared context. Every agent task lives in a Discord thread anyone on the team can see.

MCP Integration

Native support for Model Context Protocol. Connect any MCP server and expose tools to your agents in seconds.

Why nott

Built for agents. Not for generic containers.

Deploy on Render

  1. Create account
  2. New web service
  3. Connect GitHub
  4. Pick instance size
  5. Set environment variables
  6. Configure build
  7. Deploy
  8. Wait 3–10 minutes
  9. Debug if failed
  10. Redeploy

Cost for 3-person team, 5 agents: ~$149/mo

Deploy on nott

  1. @nott deploy <repo>

Cost: €49/mo. 10 seconds.

Trusted by builders

Built for engineers who ship.

Agents deployed
~10s Avg deploy time
0 Config files required

I went from 'I have an agent' to 'my agent is live with a public URL' in under 10 seconds. Zero config files.

Beta tester AI agent builder

This is what Heroku should have become for AI agents. One Discord message, everything handled.

Pricing

Pay per agent, not per container.

BYOK on every tier. Bring your own API keys — even on free.

Free

Experimenta
€0 /forever
  • 1 agent deployed
  • 1 concurrent slot
  • Haiku model (or BYOK any model)
  • 512MB storage
  • Deploy from Discord
  • No credit card. No expiration.
Deploy Your First Agent

Starter

Side Projects
€19 /per month
  • 3 agents deployed
  • 2 concurrent slots
  • Sonnet model (or BYOK)
  • 2GB storage
  • Full dashboard + telemetry
  • IDE + bash access
Get Started

Scale

Fale conosco
Custom
  • Unlimited everything
  • SLA + dedicated support
  • Custom compute resources
  • Unlimited team seats
Talk to Us

Need more?

+1 app €5/mo · +1 slot €7/mo · +1 seat €9/mo

FAQ

Questions you're probably asking.

What frameworks do you support?

Claude Agent SDK, LangChain, CrewAI, AutoGen, and any Python/Node.js agent. We auto-detect your framework from the repo.

What if my repo is private?

One-click GitHub authorization. We request access to only the repo you deploy — not your entire account.

Can I migrate from Railway/Render?

@nott deploy <your-existing-repo> — same repo, new platform. Environment variables: @nott env set KEY=value. No data migration needed.

What happens when I hit the free tier limit?

Your agent gets throttled, not killed. Add your own API key (BYOK) to remove token limits for free, or upgrade to Starter for more agents and compute.

Is there vendor lock-in?

Zero. Your code is in your GitHub repo. Your API keys are yours. If you leave, you take everything.

Do I need to know Docker?

No. You never see a Dockerfile. You never write YAML. You never touch infrastructure. That's our job.