Google released Gemini Decision 1 (gemini-decision-1) on November 11 through the Decisions API in the Gemini API. Developers can get an API key in Google AI Studio and prototype on the free tier. Paid requests cost $0.10 per million input tokens and incur no output-token charges. The 262,144-token context window covers the shared text, images, audio, video, or PDFs and up to 1,000 questions.
The model returns probabilities and confidence scores for yes-or-no checks, multiple-choice questions, and ratings against a rubric. An application can use those scores to route requests or send uncertain cases for human review. Google reports median end-to-end times of 150 to 250 milliseconds in tests with contexts up to 16,000 tokens, with little change between one and 100 questions.
