Google has introduced the Gemini Decisions API, a new programming interface designed to help AI agents make pre-defined decisions instantly without generating long text responses. The service is based on the new Gemini Decision 1 model.
Google has introduced the Gemini Decisions API, a new programming interface designed to help AI agents make pre-defined decisions instantly without generating long text responses. The service is based on the new Gemini Decision 1 model.
The main feature of the tool is that it is capable of evaluating up to 1,000 independent questions to a single multimodal input (which can contain text, images, audio, video, and PDF files) in a single pass, reports Interesting Engineering.
Instead of generating text token by token, the model evaluates ready-made variants and returns probabilities and confidence scores. This results in a latency of only 150–250 milliseconds from start to finish (with a median computation time of 50–100 milliseconds for the model itself).
Developers can use the API for limited tasks, including:
selecting tools and routing requests to the required models;
checks for compliance with rules and policies;
context assessments;
simple yes/no checks, multiple choice questions (up to 255 options), or evaluations based on predefined criteria.
The model supports a context window of up to 256,000 tokens and processes multiple data types simultaneously without the need for additional third-party processors.
Google claims that Gemini Decision 1 is about five times cheaper than the Gemini 3.5 Flash-Lite model when using the standard Interactions API.
The Decisions API costs $0.10 per million input tokens, with no fees for output tokens. Users pay only once for shared inputs, regardless of how many questions are asked. A free tier is available for prototyping in Google AI Studio.
The solution appeared against the backdrop of growing competition: in October, OpenAI introduced its own similar product ( Decisions API based on the GPT-6 Luna model).
Google is actively integrating the new API into its own products to improve their performance.
Gemini Notebook: The API is used to route incoming requests to the appropriate model layers, reducing latency for everyday tasks by 40%.
Stitch platform: helps quickly process voice and touch commands to create interactive wireframes.
CC Assistant: Instantly selects emoji reactions from over 200 options.
Google Search: serves as a protective layer to detect abuse when working with Model Context Protocol (MCP) tools.
The Decisions API is currently available through Google AI Studio, and is also integrated into popular developer platforms including OpenRouter, LangChain, Vercel, LiteLLM, and Pydantic AI.
