Google Launches Gemini Decisions API for Fast Multimodal Classification

New endpoint scores up to 1,000 yes/no, choice, and rubric questions in a single pass.

IT Solution Journal News Desk

Editorial illustration: A white square card on a dark navy background covered in a uniform grid of dots, the left two-thirds filled with solid amber circles and the right third rendered as empty blue outline circles, with a bright horizontal white light beam cutting diagonally across the full width of the card at mid-height.

AI-generated editorial illustration.

Google DeepMind announced the Gemini Decisions API on November 11, 2026, powered by the Gemini Decision 1 model. The API processes text, images, audio, video, and PDFs up to 256K tokens, returning confidence scores in 150-250 ms end-to-end latency. Google says the model scores options without token-by-token decoding, making it roughly 5x cheaper than Gemini 3.5 Flash-Lite for comparable tasks, priced at $0.10 per 1M input tokens with no output-token charges. Intended use cases include model routing, policy checks, and content grading. Google reports internal teams are already using it for query routing, emoji selection, and MCP tool safety checks. Access is available now through Google AI Studio.

Source: Google News. Original announcement: November 11, 2026.