Docs / Integrations / Claude Code

Claude Code with RelayRouter

Claude Code talks to any endpoint that implements the Anthropic Messages API. RelayRouter does, at https://relayrouter.io/v1/messages, so the whole setup is two environment variables: the base URL and your key. Billing is per token from your prepaid balance, failed requests are not charged.

#1. Create a key for a Claude group

In the console create an API key and choose a group that lists the Claude models you want. The models page shows every group with its rate; claude-opus-5-5 and claude-sonnet-4-6 are the usual picks for coding.

#2. Set the environment variables

macOS or Linux (bash, zsh):

export ANTHROPIC_BASE_URL="https://relayrouter.io"
export ANTHROPIC_AUTH_TOKEN="sk-your-relayrouter-key"
export ANTHROPIC_MODEL="claude-opus-5-5"
export ANTHROPIC_SMALL_FAST_MODEL="claude-haiku-4-5"
claude

Windows PowerShell:

$env:ANTHROPIC_BASE_URL = "https://relayrouter.io"
$env:ANTHROPIC_AUTH_TOKEN = "sk-your-relayrouter-key"
$env:ANTHROPIC_MODEL = "claude-opus-5-5"
$env:ANTHROPIC_SMALL_FAST_MODEL = "claude-haiku-4-5"
claude

To make it permanent, put the same values in the env block of ~/.claude/settings.json (or the project's .claude/settings.json):

{
  "env": {
    "ANTHROPIC_BASE_URL": "https://relayrouter.io",
    "ANTHROPIC_AUTH_TOKEN": "sk-your-relayrouter-key",
    "ANTHROPIC_MODEL": "claude-opus-5-5",
    "ANTHROPIC_SMALL_FAST_MODEL": "claude-haiku-4-5"
  }
}

#3. Pick the models

Any model id from your group works; the ids are exactly the ones shown on the models page. Inside a session /model switches the main model too.

#4. Verify outside Claude Code

If anything fails, test the same key and model with one request first:

curl https://relayrouter.io/v1/messages \
  -H "x-api-key: sk-your-relayrouter-key" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-opus-5-5","max_tokens":64,"messages":[{"role":"user","content":"ping"}]}'

#Troubleshooting