Set up Gemini

2 min read

Gemini connects with a Google AI Studio API key. Usage is billed per token to your Google account, and the connection is created in Settings > Models > Gemini.

Connect your key

Create the key

In Google AI Studio, create an API key in a project with billing enabled. Without billing, Google applies a very low usage limit and your agents stop answering quickly.

Save it as a connection

In Kujira, click Add connection under Gemini, give it a name that says which project it belongs to, paste the key, and save. Use the Test button on the connection to check that Google accepts it.

No subscription

A Google AI Pro or Ultra subscription doesn't work here: Google doesn't allow it to be used from third-party applications. Only the API key works.

Turn on an agent

In the agent, open Settings > Model, pick the Gemini connection under Connection or leave Inherit from the organization, then select a Gemini model. The change applies on its next reply, with no restart.

What to expect

• Gemini models have a one-million-token context window, so a conversation takes much longer to compact than with other providers.
• What Google counts as the model's reasoning is billed as output, and Kujira adds it to the agent's usage.
• The cost you see in Kujira is an estimate using Google's published rates. The final figure is the one in your Google Cloud console.