Models per API key
Choose the models available to each application. Different keys can have different model access.
Access,
under your control.
Choose the models and knowledge bases each application can use. Define what every person on your team can do.
01 / Application access
The right access.
For each application.
Give every API key its own scope, from the models it calls to the knowledge it can use.
Your knowledge, selectively connected.
Assign knowledge bases to the keys that need them. Each application gets access to the context you choose.
Bring context with a header.
Pass an allowed knowledge-base ID in an API header to add your own knowledge to a request.
02 / People & permissions
Roles that fit
the work.
Define what people can do in your workspace, with custom roles and individual permissions.
Build your own roles.
Choose the permissions each role includes, then assign it to your team. A reviewer might need to inspect requests and sessions; you decide what their role allows.
Make the exception. Keep the role.
Grant or deny an individual permission when someone needs different access. The exception applies to that person; everyone else keeps the role’s defaults.
03 / Usage & spend
Access has limits.
You set them.
Configure request and cost limits alongside your access controls.
Request limits
Set a boundary around request volume.
Cost limits
Define how much AI usage can cost.
Access & controls / QordIQ