Quickstart — onboard to ZetaAPI in 30 seconds
From sign-up to your first call in three steps. Covers sign-in, getting an API key, and switching to ZetaAPI in Cursor / Claude Code / any OpenAI SDK.
Updated 2026-10-05
What you’ll get in 30 seconds
- An API key that can call Claude / GPT / Gemini and other mainstream models
- ZetaAPI set as the upstream in your usual tools (Cursor / Claude Code / OpenAI Python SDK)
- Real-time token usage and cost in the console
Step 1: Sign up & top up
- Open the console and sign in with Google / GitHub / email
- Your account is created automatically on first sign-in — no review beyond email
- On the “Top up” page, add credit from $1 via card / Alipay / WeChat (a guide walks you through the first one)
Step 2: Create an API key
Go to the API Keys page in the console:
- Click
Create key - Give it a recognizable name (e.g.
cursor-mac,claude-code-work) - Copy the generated key (format
sk-xxxxx..., shown in plaintext only once)
⚠️ The full key is shown only at creation. Save it immediately to 1Password / macOS Keychain / a
.envfile or other secure place.
Step 3: Switch the upstream
Option A: Cursor
Settings → Models → OpenAI API Key:
- API Key: paste
sk-xxxxx - Base URL (enable
Override OpenAI Base URL):https://api.staging.cosmos-router.com/v1
After saving, Chat and Cmd+K all route through ZetaAPI. See the Cursor setup guide.
Option B: Claude Code
Add to ~/.zshrc or ~/.bashrc:
export ANTHROPIC_BASE_URL="https://api.staging.cosmos-router.com"
export ANTHROPIC_API_KEY="sk-xxxxx"
export CLAUDE_CODE_ENABLE_FINE_GRAINED_TOOL_STREAMING=1
After source, just run claude. See the Claude Code setup guide.
Option C: any OpenAI SDK
from openai import OpenAI
client = OpenAI(
api_key="sk-xxxxx",
base_url="https://api.staging.cosmos-router.com/v1",
)
resp = client.chat.completions.create(
model="claude-sonnet-4-6",
messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)
The official Node.js SDK / LangChain / LlamaIndex / any client that supports a custom base_url works the same way.
Option D: Claude Desktop
In the Claude Desktop menu bar, open Developer → Configure Third-Party Inference…, choose Gateway, set the Base URL to https://api.staging.cosmos-router.com and the API key to sk-xxxxx, then save and restart to use Claude models in Cowork and Code. You need to enable Developer Mode first; see the Claude Desktop setup guide for the full steps.
Verify it worked
After your first call, go back to the console home and you should see:
- Today’s usage change from 0 to non-zero
- Recent calls showing your request (model, tokens, latency, status code)
- If there’s no update within 5 minutes, your key or base URL is misconfigured — see the FAQ
Next steps
- Per-tool setup → tool setup guides
- Understand pay-as-you-go + Prompt Cache → billing
- Endpoints, error codes, rate limits → API Reference