Story

On air. The free wire runs 15 minutes behind live. See pricing

World, crypto, finance and tech news, scored for impact.

CryptoTechFlowChinese, translatedFiled 5 days ago

Google Gemini Decision 1 released: no word-by-word generation, judges 1,000 questions in a single request.

Original headline谷歌 Gemini Decision 1 发布:不做逐字生成,一次请求判 1000 个问题

Open the sourceReported by TechFlow, techflowpost.com, 12 November 2026

Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.

Google released Gemini Decision 1 and the Gemini Decisions API, a model built for closed-set decision tasks rather than token-by-token text generation, scoring up to 1,000 questions per request across text, image, audio, video and PDF inputs with 150–250ms end-to-end latency. It supports yes/no judgments, multiple choice with 2–255 options, and ordered rubric scoring, priced at $0.10 per million input tokens with no output token charge—roughly one-fifth the cost of Gemini 3.5 Flash-Lite. Google says the model leads 7 of 9 published benchmarks, including 92.2% accuracy on JevBench versus 90.9% for OpenAI's GPT-6 Luna and 90.0% for TypeSafe Jev 1.14, with the lowest calibration error, though competitors beat it on two tests and total cost. The launch matters because it targets AI routing, agent tool selection, product design and safety detection, positioning the model as a low-cost judgment layer inside AI systems rather than a chatbot. Google has already deployed it internally; next to watch is developer adoption through Google AI Studio and whether the pricing and latency hold up at production scale.

Muhabbit's AI wrote this summary in its own words from TechFlow's report and translated it from Chinese. It can get details wrong, so check the source.

Named in this story

Google (GOOGL), Gemini Decision 1, Gemini Decisions API, Sundar Pichai, Google AI Studio

A linked name opens that asset's sentiment page.