UniRoute

UniRoute integrations

One API key for coding tools, chat clients and creative tools — each with a setup guide.

Set up with one command

npx -y uniroute-setup@latest

Prompts for your key, writes the client config and keeps a backup.

  • Claude Code
  • Codex
  • Gemini CLI
  • OpenCode
  • OpenClaw
  • Hermes Agent
Create an API keyLet your AI agent set it up

All tools18

Terminal agents6

Claude Code

One command

Set the gateway URL, API key and Claude model in your user settings.

settings.json · ANTHROPIC_BASE_URL

Codex

One command

Add the UniRoute provider, choose a model and configure API key authentication.

config.toml · Custom provider

Gemini CLI

One command

Set the native Gemini URL and key, then select a Gemini model.

.env · GOOGLE_GEMINI_BASE_URL

OpenCode

One command

Generate a provider for the model’s protocol and set your default model.

opencode.jsonc · Native providers

OpenClaw

One command

Configure a provider with the setup tool or sign in through the UniRoute plugin.

openclaw.json · Providers & key

Hermes Agent

One command

Add a UniRoute provider and select Messages, Responses or Chat Completions.

config.yaml · Custom provider

Editors & IDEs5

Cline

Open API Configuration, choose OpenAI Compatible, then enter the UniRoute base URL, key and model ID.

VS Code settings · custom Base URL

Kilo Code

Add a Custom provider, select OpenAI Compatible, then enter the UniRoute URL, key and model ID.

VS Code Providers · add a custom endpoint

Cursor

Open Models, save your UniRoute key under OpenAI API Key, enable Override OpenAI Base URL, and select your custom model.

Models · custom OpenAI base URL

JetBrains IDEs

In Providers & API keys, choose OpenAI-compatible, enter the UniRoute URL and key, test the connection, then assign the model.

AI Assistant · OpenAI-compatible provider

TRAE (international)

Add a Custom Model, choose Anthropic Messages, enter the UniRoute URL and key, then turn off Auto Mode and select the model.

Models · custom API format and URL

Desktop apps2

Claude desktop

Enable Developer Mode, enter the gateway URL and key, then add model IDs.

Developer menu · Third-party inference

Codex desktop

One command

Close the app, configure the Codex provider and file-based key, then reopen it.

config.toml + auth.json · Shared with CLI

Chat & writing clients2

Cherry Studio

Add UniRoute, fill the OpenAI endpoint and API Key, then add a model and enable the provider.

Model Provider · protocol endpoints

Chatbox

Add an OpenAI-compatible provider, enter the API Host, API Path and key, then add a model and click Check.

OpenAI compatible · host + path

Video & creative tools2

MoneyPrinterTurbo

Coming soon

Setup guide coming soon.

Video creation · API integration

HyperFrames

Coming soon

Setup guide coming soon.

Video creation · API integration

Config managers1

CC Switch

Add a custom provider, enter the URL and key, then fetch and map models.

Provider settings · Model mappings

Tool not listed?

Any client that accepts an OpenAI-compatible or Anthropic base URL can connect.

https://api.uniroute.app/v1(OpenAI compatible)
https://api.uniroute.app(Anthropic / Gemini)
API formats and addresses

Frequently asked questions

Can I use the same key across integrations?

Yes. Use the same UniRoute API key across tools, or create separate keys to track each tool’s usage.

Which API formats does each model support?

Each model’s page lists its supported API formats. Match the model’s format to your client, then use the address in the formats guide. API formats and addresses →

Does setup send a model request?

No. Setup writes configuration and queries the model catalog. It does not send a model request. Requests you send from the client are billed by usage.

What if my tool is not listed?

Add a custom provider with your UniRoute key and the base URL for its API format. The formats and addresses guide shows the endpoints. API formats and addresses →