OpenClaw

OpenClaw can be configured with a LiteLLM-style custom base URL. Use R9S as the base URL and your R9S API key as the provider key.

Environment Variables

export R9S_API_KEY="<R9S_API_KEY>"
export LITELLM_API_KEY="$R9S_API_KEY"

Onboard With R9S

Use the LiteLLM-compatible auth flow and set the custom base URL to R9S:

openclaw onboard \
  --non-interactive \
  --accept-risk \
  --auth-choice litellm-api-key \
  --litellm-api-key "$R9S_API_KEY" \
  --custom-base-url "https://api.r9s.ai/" \
  --install-daemon \
  --skip-channels \
  --skip-skills

Then select a model routed through R9S:

openclaw models set litellm/claude-sonnet-4.5

Run OpenClaw

openclaw agent --agent main -m "Hello, what model are you?"

You can use another R9S model by replacing the model suffix:

openclaw models set litellm/gpt-5-codex

Troubleshooting

Symptom	Check
Authentication error	Confirm LITELLM_API_KEY or the onboarding key is your R9S API key.
Model not found	Confirm the model name exists in R9S, then use litellm/<model-name> in OpenClaw.
Requests go to the old proxy	Re-run onboarding or update the saved custom base URL to https://api.r9s.ai/.