Google 发布 Gemini Decisions API 与 Gemini Decision 1 模型
Google 把模型从写文本改成了做选择题，一次能判 1,000 个问题，输出 token 不收钱，比 Flash-Lite 便宜约 5 倍。做内容审核、自动打分、批量评测的工程师可以直接去 AI Studio 试试，成本结构跟生成类 API 完全不同。
Google 推出 Gemini Decisions API，底层由新模型 Gemini Decision 1 驱动，已在 Google AI Studio 开放使用。与逐 token 生成文本的传统模型不同，该模型专为封闭集选择题设计，单次调用最多可对 1,000 个问题打分，支持文本、图像、音频、视频和 PDF 多种输入。
模型提供三种判定类型：是非题、最多 255 个选项的多选题，以及基于评分标准（rubric）的打分。定价上输入 token 为 $0.10/1M，输出 token 完全免费，官方称整体比 Flash-Lite 便宜约 5 倍，并在多个基准评测中领先，校准误差最低。
这个产品思路是把 LLM 从生成器改造成低成本分类器，适合内容审核、表单判定、评测打分这类高频批量场景。演示中 Google 内部团队已用它在 Stitch Rapid 里几秒内完成完整产品流程设计。对需要大规模判定任务的开发者来说，输出零计费的定价结构会直接影响成本模型。
Today we’re launching the Gemini Decisions API, powered by our new Gemini Decision 1 model in @GoogleAIStudio . Instead of generating text token by token, it’s built for closed-set choices scoring up to 1,000 questions across text, images, audio, video + PDFs in a single pass. A few details: - 3 decision types: yes/no, multiple choice (up to 255 options) + rubric scoring - $0.10 / 1M input tokens with $0 billed for output tokens (~5x cheaper than Flash-Lite) - leads on several benchmark evals with lowest calibration error Google teams are already building with it, below is a look at Stitch Rapid designing a full product flow in seconds (1x speed)
This report is categorized under AITOP's "Products" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Gemini Decision 1, Google AI Studio, Decisions API, 多模态.
Source attribution: curated from Sundar Pichai (@sundarpichai).
