Model API Docs
How onboarding, authentication and rate limits work for the Model API.
Availability
The Model API is available on request to enterprise customers. Technical reference material is shared with each customer during onboarding.
Onboarding
- Enquiry. Contact us through with your organisation and use case.
- Scoping. We agree the models, expected volumes and rate limits you need.
- Agreement. Access is provided under a written agreement.
- Credentials. We issue credentials to your nominated technical contacts through a private channel.
- Integration. We share the reference material and support your team through the first integration.
Authentication
Each customer receives its own credentials. Requests are authenticated with those credentials and must be sent over encrypted connections.
- Keep credentials on your servers; never put them in browser or mobile code.
- Credentials can be rotated or revoked on request, and immediately if you suspect exposure.
- Access is scoped to what your agreement covers.
Rate limits
Rate limits are agreed per customer. Requests above your agreed limits are refused until capacity is available again, so build retries with back-off into your integration.
Data handling
Requests pass through the same controlled gateway as our products. Data handling terms are set out in your agreement and are consistent with our Privacy Policy.
Support and changes
Your nominated contacts can reach us through Enquiries. We give customers notice before changes that affect their integration.