Google 推出 Gemini Decision 1 决策模型与 APIChinese only
Google 推出 Gemini Decision 1 和 Decisions API，用于一次性判断是非题、多选题及评分任务；每次请求最多处理 1,000 个问题，并为每个答案附置信度分数。Stitch 已用该 API 试做从简要提示生成多屏产品流程的功能，LangChain 也推出了首日支持。
1 developments4 reports4 sources3 verified facts
Verified facts 3
谷歌推出 Gemini Decisions API 快速多模态决策，基于 Gemini Decision 1 模型
谷歌宣布推出由 Gemini Decision 1 模型驱动的 Gemini Decisions API，可对多模态内容进行批量决策，且该 API 已在 Google AI Studio 提供。
Gemini Decisions API: Fast, calibrated multimodal decisions
谷歌推出由Gemini Decision 1模型驱动的Gemini Decisions API，可通过Google AI Studio中的Gemini API使用，用于对多模态输入进行快速的限定选项判断。
Gemini Decisions API
谷歌推出 Gemini Decision 1 决策模型 API，支持是非、选择和评分问题，并可处理文本、图像、音频、视频及 PDF。
Event developments 1
Google DeepMind launches Gemini Decisions API for multiple multimodal decisions in one pass
The Gemini Decisions API is available through the Gemini API in Google AI Studio. It uses Gemini Decision 1 to answer up to 1,000 yes/no, multiple-choice or scoring questions about a shared input containing text, images, audio, video or PDFs in one pass. It costs $0.10 per million input tokens, has no output-token charges and offers a no-cost tier for prototyping.
View source reports · 4 sources · 4 reports
- Google AI 官方博客 · Google DeepMind launches Gemini Decisions API for multiple multimodal decisions in one pass
- Simon Willison AI 实践 · Google introduces the Gemini Decision 1 decision-model API
- IT之家 科技新闻 · Google launches Gemini Decisions API for up to 1,000 multimodal decisions per request
- Latent Space AI 报道 · Google launches Gemini Decision 1 and Decisions API
All reports 4
Google launches Gemini Decision 1 and Decisions API
Google launched Gemini Decision 1 and the Decisions API for single-pass yes/no, multiple-choice and rubric-scoring tasks. It handles up to 1,000 questions per request and attaches a confidence score to each answer. Stitch has built a prompt-to-product-flow experiment with the API, while LangChain has shipped day-zero support.
3 more sources reported this
- IT之家 科技新闻
- Simon Willison AI 实践
- Google AI 官方博客
Google launches Gemini Decisions API for up to 1,000 multimodal decisions per request
Powered by Gemini Decision 1, the API lets developers ask up to 1,000 independent questions about a single input containing text, images, audio, video or PDFs. It is designed for tasks including tool selection, policy checks and model routing. The API is available in Google AI Studio with a free allowance for prototyping; input costs $0.10 per million tokens, while output tokens are free.
3 more sources reported this
- Latent Space AI
- Simon Willison AI 实践
- Google AI 官方博客
Google introduces the Gemini Decision 1 decision-model API
Gemini Decision 1 supports yes/no questions, choices and rubric scoring across text, images, audio, video and PDFs. Input costs 10 cents per million tokens, with free output. Among the compared services, Jev handles only text and JSON, while OpenAI handles text and images.
3 more sources reported this
- Latent Space AI
- IT之家 科技新闻
- Google AI 官方博客
Google DeepMind launches Gemini Decisions API for multiple multimodal decisions in one pass
The Gemini Decisions API is available through the Gemini API in Google AI Studio. It uses Gemini Decision 1 to answer up to 1,000 yes/no, multiple-choice or scoring questions about a shared input containing text, images, audio, video or PDFs in one pass. It costs $0.10 per million input tokens, has no output-token charges and offers a no-cost tier for prototyping.
3 more sources reported this
- Latent Space AI
- IT之家 科技新闻
- Simon Willison AI 实践