Gemini Decisions API: 1,000 questions at $0.10 per 1M input

Google said on November 11, 2026 that it is launching the Gemini Decisions API, powered by a new model, Gemini Decision 1 (gemini-decision-1), in the Gemini API in Google AI Studio. The company says the endpoint can evaluate up to 1,000 yes/no, multiple-choice and rubric-scoring questions against one multimodal input in a single pass, at $0.10 per 1 million input tokens, with zero output tokens billed.

Logan Kilpatrick, writing on X the same day, gave the same input price, said there is "no additional cost," and noted native support for images, audio, video and PDFs, a 256K context window, and 1,000 questions.

What it does

Rather than generating answer text token by token, Gemini Decision 1 "scores your candidate options in a single forward pass, returning probabilities and confidence scores," Google said, in 150 to 250 milliseconds end-to-end. The blog lists three question types that can be mixed in one request: yes/no checks, multiple choice (2 to 255 options), and an ordered rubric that returns a weighted score and per-level probabilities. Every answer includes a confidence score from 0 to 1.

The shared input can be text, images, audio, video and PDFs, up to 256K tokens. Google says that input is billed once per request no matter how many questions you evaluate, and that each question adds only its own instruction and criteria tokens. There are no output-token charges, "variable thinking-token bills, or separate audio transcription and OCR costs," according to the post. A no-cost tier in Google AI Studio is available for prototyping.

Google claims Decision 1 is nearly five times cheaper, on a shared benchmark suite, than Gemini 3.5 Flash-Lite with the Interactions API, and that it "leads across the suite, with the lowest calibration error." Those are Google's own comparisons, not an independent audit. The company also says teams inside Google already use the endpoint for model routing in Gemini Notebook, design iteration in Stitch, emoji selection in CC, and a safety layer for MCP tools in Search.

Ecosystem integrations are "rolling out" across OpenRouter, LangChain, Vercel, LiteLLM and other tools, Google said, "with support coming soon to Gemini in Google Cloud."

Codema's AI agents hub, AI models guide, and Gemini 4 Argon pricing article cover the broader Gemini lineup this endpoint sits beside. What to watch is whether the 150-to-250-millisecond latency and $0.10 input price hold up outside Google's own tests, and when Cloud support actually ships.