Introduction
What Model Gate is, how cost and usage cost differ, and which API formats are supported.
Introduction
Model Gate is an AI gateway with OpenAI-compatible and Anthropic-compatible APIs. One API key can use every model enabled for that account.
Workflow
1. Create an API key in the control panel. 2. Fund the Model Gate account balance. 3. Select a canonical model ID or enabled alias. 4. Send the request to https://api.model-gate.com. 5. Model Gate resolves the route and forwards the canonical model upstream. 6. The actual request cost is debited from the account. 7. The key and group usage limits are updated with the independently calculated usage cost.
Cost and usage cost
- cost is the actual balance debit. - usage_cost affects only resettable API-key and group limits. - official_base_cost is the request cost at saved official provider rates.
The effective usage rule is configured on the key or inherited from its group.
Supported API formats
| Format | Base URL | Main endpoint |
|---|---|---|
| Anthropic | https://api.model-gate.com | POST /v1/messages |
| OpenAI | https://api.model-gate.com/v1 | POST /v1/chat/completions |
| OpenAI Responses | https://api.model-gate.com/v1 | POST /v1/responses |
See the endpoint reference and connection guides.