Run your own AI gateway in 5 minutes: OmniRoute, one click, your own URL

An AI gateway fixes this. One OpenAI-compatible endpoint, many providers behind it, automatic fallback when one runs out

Every AI coding tool wants its own provider config. Every provider has its own rate limit. You hit one mid-task and start swa pping keys by hand.

An AI gateway fixes this. One OpenAI-compatible endpoint, many providers behind it, automatic fallback when one runs out.

OmniRoute (MIT, by @diegosouzapw) is one of the best open-source options:

The hard part was always hosting it. So we made it one click.

Deploy: nexusai.run/deploy?repo=https://github.com/diegosouzapw/OmniRoute

What happens

Zero config. OmniRoute generates its own secrets on first start.

Finish setup

Open your URL. The wizard asks for a setup token.
Copy it from your deployment's Logs tab and set a password.
Connect a provider, copy an API key from Endpoints.
Point Claude Code, Codex, Cursor, or any OpenAI client at https://<your-app>.nexusai.run/v1

Security

The API needs a key from the start
The dashboard stays locked until setup, using a token only you can see in your logs
It runs in your account, so your provider keys stay yours

Cost

Runs on the NEXUS AI Free plan. Pair it with free-tier providers and it costs nothing.

Note: data lives in the container. Stop and start keeps it. A full rebuild starts fresh.

For maintainers

The same Deploy to NEXUS AI button works on any public GitHub repo. One line in your README, and anyone can run your app, dat abase included.

Guide: nexusai.run/kb/deploy-button

Star OmniRoute: github.com/diegosouzapw/OmniRoute

About NEXUS AI

NEXUS AI is an agentic AI app builder and full-stack deployment platform. Explore the AI App Builder, learn more on the About page, read the documentation, or contact the team through nexusai.run/contact.