Google launches Gemini Decisions API with 150-250ms response times

Google has launched the Gemini Decisions API, powered by its Gemini Decision 1 model, enabling AI agents to make predefined decisions in 150–250 milliseconds by scoring options in a single pass. The API evaluates up to 1,000 questions against multimodal inputs like text, images, and audio. The API supports yes-or-no checks, multiple-choice questions with up to 255 options, and rubric-based scoring, with median model computation time of 50–100 milliseconds. Google claims it outperforms competitors on benchmarks and is nearly five times cheaper than Gemini 3.5 Flash-Lite, costing $0.10 per million input tokens with no output charges. The launch follows OpenAI's October Decisions API, powered by GPT-6 Luna, which supports up to 400 questions per request. Google is already using the technology in products like Gemini Notebook, Stitch, and Google Search, and it is available via Google AI Studio with integrations for OpenRouter, LangChain, and Vercel.