Google launches Gemini Decision 1 for high-speed, low-cost evaluations
Google DeepMind has introduced the Gemini Decisions API, a service designed for fast, closed-ended evaluations such as yes-or-no questions, multiple-choice decisions and scale-based ratings. Its underlying model, Gemini Decision 1, can process up to 1,000 independent questions in a single pass, using text, images, audio, video and PDFs as input. Instead of generating conventional answers, it scores the available options and returns probabilities and confidence values.
Google says responses arrive in 150–250 milliseconds, making the system suitable for high-volume automated checks where speed and predictable output matter more than open-ended reasoning. The company is already using the API for dynamic model routing in Gemini Notebook, design prototyping in Stitch and safety checks in Google Search. The service costs $0.10 per million input tokens and does not charge for output tokens, which Google says makes it about five times cheaper than Gemini 3.5 Flash-Lite.
Gemini Decisions is available through Google AI Studio, while integrations with OpenRouter, LangChain, Vercel and other platforms are being introduced. The launch also places Google in a growing market for specialized decision models, following comparable offerings from Jev, OpenAI, Cloudflare and Microsoft.
Continue in the app — vote & join in ➔