Getting started

Quickstart

1. Get your key

Sign up with email, GitHub, or Google. You'll get 100 credits free — about three million tokens of routed traffic. Your default key is on the dashboard the moment you log in.

2. Pick an API shape

Omnizen speaks two API shapes natively. Use whichever your tool already speaks — no code changes:

Anthropic Messages — for Claude Code, Cline, OpenCode, and anything else built around anthropic-version.
OpenAI Chat Completions — for Cursor, Codex CLI, Aider, Continue, and the rest.

3. Set the env vars

For a tool that speaks the Anthropic shape:

bash

export ANTHROPIC_BASE_URL="https://api.omnizen.ai"
export ANTHROPIC_AUTH_TOKEN="sk-omni-…"

For a tool that speaks the OpenAI shape:

bash

export OPENAI_BASE_URL="https://api.omnizen.ai/v1"
export OPENAI_API_KEY="sk-omni-…"

4. Send a request

A direct sanity check — paste this into a terminal with your key in the environment:

bash

curl https://api.omnizen.ai/v1/messages \
  -H "x-api-key: $ANTHROPIC_AUTH_TOKEN" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "deepseek-v4",
    "max_tokens": 256,
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

5. Pick your tool

From here, jump to the guide for the tool you actually use:

Smart routing, manual selection

By default, Omnizen routes your request to the cheapest credible frontier model for the task. Pass an explicit model and we route to that one. Switch routing modes (Smart Quick, Smart Expert, or Manual) any time from the dashboard.