Your keys. Your providers.
Steinkauz AI does not resell inference. Connect public cloud, private cloud and on-premises at the same time, through provider-specific integrations or the OpenAI-compatible connector. The same policy sits in front of them.
Public managed
OpenAI
Anthropic
Mistral- And many more
Private managed
Amazon Bedrock
Azure OpenAI
Google Vertex AI- And many more
Self-hosted
- OpenAI-compatible
For the full list and descriptions, see BYOK Providers in the docs. Missing a provider? Contact us to add it.
Bring Your Own Key (BYOK)
Connect supported public, private-cloud or on-premises providers. Supply your own credentials where the service requires them.
- Your provider relationship. No inference markup from Steinkauz AI.
- Azure OpenAI, OpenAI-compatible endpoints, Anthropic, Google and more
- Compatible local services such as TEI can use keyless access and manual model IDs when model listing is unavailable.
- Optional budgets gate recorded usage. Unreported provider usage and cost remain unknown.
Routing policy is configured under Governance. Incompatible requests are blocked before inference.
Models
Choose models from your own providers
Steinkauz AI does not sell a separate model catalogue. Choose models on your configured connections for the operations their adapters and upstream services support.
- Use supported language models, text embedding services and compatible speech transcription endpoints.
- Enter model IDs manually when a compatible service does not expose a model list.
- Choose a connection and model in your application. Policies governs D/E routing permissions.
Connect your providers
Bring your own keys. The models you can use are the models those providers already include.