Gemini Decision 1
Google's decision model that answers up to 1,000 yes/no, choice and score questions about text, images, audio, video or PDFs in one request, returning probabilities instead of text.
Google's decision model that answers up to 1,000 yes/no, choice and score questions about text, images, audio, video or PDFs in one request, returning probabilities instead of text.
Overview
Gemini Decision 1 powers Google's Decisions API. You send shared input, such as a support message, a photo or a video frame, with questions that have fixed answers: yes or no, a choice between 2 and 255 options, or a score on a rubric of 2 to 255 levels. It returns a probability for each answer and a confidence score instead of generating text, so software can route, filter or score directly.
It bills only input tokens, at $0.10 per million, and charges the shared input once however many questions you ask. Google lists request routing, policy checks across many rules, rubric scoring, tool selection and live media classification as uses. It uses the same Choice, Score and Noul question types as Cloudflare's Clef models, and also reads audio, video and PDFs, with up to 262,144 tokens per request.
Category comparison
These are provider-published specifications, not Cody benchmark scores. Follow the linked sources for current limits and endpoint-specific exceptions.
API and provider access
Availability
Available through the Decisions API on the Gemini Developer API (v1beta) and in Google AI Studio, on free and paid tiers.
Access is limited to Google's supported regions and eligible accounts.
Check live availabilityData and training
On the Gemini API's free tier, Google may use prompts and responses to improve its products; on the paid tier, it does not.
This is a concise reading of the cited provider material, not legal advice. A third-party gateway can have different storage, routing, training, and residency terms from the model maker's direct API.
Read the provider policyFrequently asked questions
Google's decision model that answers up to 1,000 yes/no, choice and score questions about text, images, audio, video or PDFs in one request, returning probabilities instead of text. Gemini Decision 1 powers Google's Decisions API. You send shared input, such as a support message, a photo or a video frame, with questions that have fixed answers: yes or no, a choice between 2 and 255 options, or a score on a rubric of 2 to 255 levels. It returns a probability for each answer and a confidence score instead of generating text, so software can route, filter or score directly.
Gemini Decision 1 was released on November 11, 2026 according to the cited provider materials.
Available through the Decisions API on the Gemini Developer API (v1beta) and in Google AI Studio, on free and paid tiers. The access routes listed in this guide are Google.
$0.10 / 1M input tokens; output free. Each request bills the shared input once, however many questions it asks; each question adds the tokens of its own instructions and options. Images, audio, video and PDFs count as input tokens. Caching and the Batch API are not available. A rate-limited free tier is also offered.
Available through the Decisions API on the Gemini Developer API (v1beta) and in Google AI Studio, on free and paid tiers. Access is limited to Google's supported regions and eligible accounts.
On the Gemini API's free tier, Google may use prompts and responses to improve its products; on the paid tier, it does not. The policy belongs to the provider route and account terms, so verify it again before production use.
Related comparisons
Cloudflare's preview extraction model that turns emails, receipts and other unstructured text and images into typed values that software can use directly.
Open full comparisonMicrosoft's decision model for sorting requests, choosing a route and scoring results without generating a chat response.
Open full comparisonCloudflare's larger decision model for classifying information and choosing an action without generating a chat response.
Open full comparisonA smaller Cloudflare decision model for frequent routing and classification tasks, with a lower input-token price than Clef.
Open full comparisonTypeSafe AI's first System One model for fast, typed choices, scores, and probabilities that software can use directly.
Open full comparison