Google Launches Gemini Decisions API Powered by Gemini Decision 1
Google has announced the Gemini Decisions API, driven by the new Gemini Decision 1 model, designed for fast multimodal decision-making tasks. It allows developers to evaluate up to 1,000 independent questions on text, images, audio, video, and PDF content in a single pass. This API significantly lowers latency and cost for specialized decision tasks such as tool selection, policy checks, and model routing. By handling bulk structured questions efficiently with low latency, it streamlines the development of advanced AI applications and agents. The API processes inputs up to 256K tokens with a latency between 150 to 250 milliseconds, charging $0.10 per million input tokens while output tokens are free. It evaluates candidate options through a single forward pass, returning probabilities, confidence scores, and support for up to 255 options.
## BACKGROUND
Multimodal AI refers to artificial intelligence systems that can simultaneously process and understand multiple types of data modalities, such as text, images, audio, and video. Model routing involves dynamically directing user prompts or tasks to the most appropriate AI model based on cost, speed, and capability requirements.