Gemini

Model name: gemini

About Gemini

Google Gemini is a family of multimodal large language models developed by Google. The models are accessible via the Gemini API.

Supported aidb operations

• decode_text
• decode_text_batch

Supported models

Any text generation model that the provider supports. See a list of supported Gemini models here.

Creating a Gemini model

You can create any supported Gemini model using the aidb.create_model function.

This example creates a Gemini 2.5 Flash model with the name my_gemini_model:

SELECT aidb.create_model(
  'my_gemini_model',
  'gemini',
  '{"model": "gemini-2.5-flash"}'::JSONB,
  '{"api_key": "AIzaSyBXnKr4yJG2lMhyvza81q3-WFuP7Tk3r8z"}'::JSONB 
);

Running the model

SELECT aidb.decode_text('my_gemini_model', 'Why is the sky blue?');

Model configuration settings

The following configuration settings are available for OpenAI models:

• model — The model to use.
• url — The URL of the model to use. This setting is optional and can be used to specify a custom model URL.
• max_concurrent_requests — The maximum number of concurrent requests to make to the OpenAI model. The default is 25.
• thinking_budget — The number of thinking tokens used to generate responses. This setting is optional. See more here.

Model credentials

The following credentials may be required by the service providing these models.

• api_key — The API key to use for Bearer Token authentication. The api_key is sent in a header field as Authorization: Bearer <api_key>.