On 11 November 2026 Google introduced the Gemini Decisions API, a new endpoint for fast, closed-set decisions: picking a tool, checking a policy, routing a request to the right model or grading context. It runs on a new model, Gemini Decision 1 (gemini-decision-1), which does not generate an answer word by word: it scores the candidate options in a single forward pass and returns probabilities.
How it works: one request carries one shared input (text, images, audio, video and PDFs, up to 256K tokens) and up to 1,000 independent questions of three kinds: yes/no, multiple choice (2 to 255 options) and rubric scoring. Every answer comes with a confidence score from 0 to 1, so your code can send uncertain cases to a fallback model or a person. Google gives 150–250 ms end-to-end latency and says it stays nearly flat from 1 to 100 questions per request.
Price: $0.10 per million input tokens and $0 for output; the shared input is billed once per request, however many questions you ask. Google says it is nearly 5 times cheaper than Gemini 3.5 Flash-Lite on its own benchmark suite and that the model leads those benchmarks with the lowest calibration error, but the post gives no numeric scores, so these are Google's claims, not independent tests. Google already uses it as a router in Gemini Notebook (up to 40% lower latency, it says) and as a safety layer for MCP tools in Search.
How to try it: get a key in Google AI Studio, which has a no-cost tier for prototyping, and call client.decisions.create in the google-genai SDK (Python, JavaScript or REST). Developer Flavio Copes notes it needs @google/genai 3.2 or later on npm and lists free-tier limits of 200 requests per minute and 10,000 per day; he also points out that context caching, the Batch API and conversation history are not supported. Google says support in Gemini in Google Cloud is "coming soon" and does not say whether the API is in preview or generally available.
The launch reached the Hacker News front page within hours (about 180 points). Integrations are listed for OpenRouter, LangChain, Vercel, LiteLLM, Pydantic AI and others.
