Summary

The Gemini Decisions API is a purpose-built Google endpoint launched in November 2026 for fast, closed-set decisions, powered by the new model Gemini Decision 1. Instead of generating text, the model scores candidate answers in a single forward pass and returns calibrated probabilities with confidence scores, in 150 to 250 milliseconds end to end. One request can evaluate up to 1,000 independent questions against a single shared multimodal input, which is billed once no matter how many questions are asked.

Key Features

Up to 1,000 independent questions evaluated in a single request and forward pass
Three question types: yes/no, single-select choice with 2 to 255 options, and rubric score with 2 to 255 levels
Every answer includes a confidence score from 0 to 1 for fallback and human routing
Zero output tokens, so output is never billed
Native multimodal input with text, images, audio, video and PDFs in a 262,144-token shared context
150 to 250 ms end-to-end latency, nearly flat from 1 to 100 questions
Shared input billed once no matter how many questions are asked, at $0.10 per 1M input tokens
Integrations with OpenRouter, LangChain, Vercel AI SDK, LiteLLM, Pydantic AI, Pipecat, LiveKit and OpenCode

About Gemini Decisions API


Developers send one shared input of text, images, audio, video or PDFs with up to a 262,144-token context, plus typed questions: yes/no, single-select choice with 2 to 255 options, or rubric score with 2 to 255 ordered levels. Every answer carries a confidence score from 0 to 1 for fallback or human routing, and because nothing is generated there are zero output tokens and output is never billed. The API is called with client.decisions.create in google-genai SDK 3.2 or newer for Python and JavaScript, or over REST with a Gemini API key from AI Studio.


Typical uses are model routing, policy checks, tool selection, context grading and classification, where code branches on the returned probabilities. Pricing is $0.10 per 1M input tokens with $0 output tokens, and a no-cost tier in Google AI Studio covers prototyping. Integrations are rolling out with OpenRouter, LangChain, the Vercel AI SDK, LiteLLM, Pydantic AI, Pipecat, LiveKit and OpenCode.

Get more likes & reach the top of search results by adding this button on your site!

Embed button preview - Light theme
Embed button preview - Dark theme
TurboType Banner

Subscribe to the AI Search Newsletter

Get top updates in AI to your inbox every weekend. It's free!