Sakhanda Wire
NVDA $153.77 -1.52% MSFT $504.91 -2.53% GOOGL $281.52 -0.78% META $620.99 -0.87% AMZN $256.80 -0.37%
← Back to the news

Gemini Decisions API brings fast multimodal scoring

Google has introduced the Gemini Decisions API, a new endpoint for bounded decisions powered by the Gemini Decision 1 model. Available through the Gemini API in Google AI Studio, it can assess up to 1,000 yes/no, multiple-choice, or rubric-scoring questions against one shared input containing text, images, audio, video, and PDFs, with contexts of up to 256K tokens.

Instead of generating responses token by token, the model scores candidate options in a single pass. The API returns probabilities and confidence values, with reported end-to-end latency of 150–250 milliseconds and a price of $0.10 per 1M input tokens. Output tokens are not billed, and the shared input is charged once per request regardless of the number of questions.

Google says Gemini Decision 1 achieved the lowest calibration error across benchmarks covering agent routing, reasoning, ranking, and multimodal understanding. The company describes uses including model routing in Gemini Notebook, rapid wireframe edits in Stitch, emoji selection in CC, and a safety layer for MCP tools in Search. A no-cost tier is available in Google AI Studio, while integrations are being introduced for several developer platforms.

Written by Sakhanda Wire AI from reporting by Google AI. Facts, quotes and full context are in the original article below.

Originally published by Google AI on

Read the original on Google AI ↗

The reporting and the image belong to Google AI and are credited and linked above; the text on this page is Sakhanda Wire's own summary.

← Back to the news

More stories

All the latest news