GenAI Gateway

The GenAI Gateway connects Knime workflows to large language models (LLMs) and embedding models from external providers, such as OpenAI or Azure OpenAI. Knime doesn't host these models. The gateway manages the connection to the providers you configure.

With the gateway, admins register a model once and share it with teams, instead of distributing provider credentials to each user. To learn how the gateway routes requests and how model scope and access work, see How GenAI Gateway works.

Activate the GenAI Gateway

The GenAI Gateway is an optional feature and isn't enabled on your instance by default.

To activate it:

1. Contact Knime support with a request to enable the GenAI Gateway.
2. Provide your AI provider connection details, such as API keys and endpoint URLs. For the supported platforms, see Supported platforms.
3. Knime configures the gateway on your instance.

Manage models

Once the gateway is active, admins create, edit, and share models:

• Global admins manage global GenAI models from the Global administration page.
• Team admins manage team GenAI models from the team's page.

For the steps, see Manage and use AI models. To configure the AI providers that K-AI uses, see Bring your own model.