Models and Capabilities
OpenFrog sits in front of multiple upstream providers and exposes them through one gateway layer for applications and agent workflows.
Typical usage patterns
- Use standard OpenAI-style interfaces for chat, embeddings, or image features
- Assign models, routes, and quota from one admin console
- Reuse the same key and Base URL across multiple agents or clients
Confirm these items first
- The model is enabled for your account
- The upstream route is configured in the console
- The client model name matches the name exposed by the gateway
Recommended rollout approach
For most teams, the shortest successful path matters more than full coverage on day one:
- Get a key
- Set the Base URL
- Choose one enabled model
- Complete the first request