Google has launched the Gemini Decisions API, a new endpoint powered by its Gemini Decision1 model that helps AI agents make rapid, predefined decisions without generating lengthy text responses.
This API enables developers to evaluate up to 1,000 independent questions against shared multimodal inputs containing text, images, audio, video, and PDFs. Gemini Decision1 does not generate responses token by token; instead, it scores predefined options in a single forward pass, returning probabilities and confidence scores with an end-to-end latency of 150 to 250 milliseconds.
Google states that the model is designed for bounded tasks, such as selecting tools, checking policy compliance, routing queries to the appropriate model, and evaluating context. Developers can submit "yes or no" checks, multiple-choice questions with up to 255 options, or scoring requests based on predefined criteria in a single API call.
Faster, lower-cost decisions
According to Google, the median model computation time for this API is 50 to 100 milliseconds, and performance remains nearly unchanged when processing 1 to 100 questions per request.
This model supports a context window of up to 256,000 tokens and can process multiple types of data simultaneously, enabling developers to combine inputs such as audio recordings, images, and documents without relying on separate processing systems.
Google also claims that the model outperforms competing systems on benchmarks covering agent routing, reasoning, ranking, and multimodal understanding, while achieving the lowest calibration error.
The company stated that, in its shared benchmark suite, Gemini Decision1 is nearly five times cheaper than Gemini 3.5 Flash-Lite using the Interactions API.
This release comes amid intensifying competition in the professional AI decision-making tools market. OpenAI launched its Decisions API in October, powered by its GPT-6 Luna model. The company states that this service is 10 times faster than its Responses API and supports up to 400 questions per request. Similar to Google’s offering, it provides probabilistic checks, multiple-choice selection, and scoring based on rubrics, with pricing starting at $0.10 per million input tokens.
The Decisions API charges $0.10 per million input tokens and does not charge for output tokens. Developers pay only once for shared input data, regardless of how many questions are evaluated, even though each question adds tokens related to instructions and criteria.
Empowering AI applications in the real world
Google has deployed this technology across multiple products, demonstrating how specialized decision models can enhance the speed and efficiency of automated systems.
In Gemini Notebook, the API routes incoming queries to the appropriate model tier and inference level. According to Google, this approach can reduce latency for routine queries by up to 40%, while reserving more powerful reasoning models for complex tasks.
This technology also supports rapid design experiments on Google's Stitch platform, enabling users to generate interactive wireframes and make changes via voice commands or touch input.
Meanwhile, Google's CC Assistant uses this API to select emoji reactions from over 200 options in a fraction of a second.
Google Search also uses this technology as a security layer to detect adversarial behavior involving Model Context Protocol (MCP) tools.
In addition to these applications, this API can help developers build systems that automatically select appropriate tools, evaluate incoming information, and determine when human intervention is needed.
The Decisions API is now available through Google AI Studio, including a free tier for prototyping. Google is also rolling out integrations with developer platforms such as OpenRouter, LangChain, Vercel, LiteLLM, and PydanticAI.


