11-19 更新 实时
首页/ 技术栈/ 正文

Google 推出 Gemini Decisions API

技术栈 来源:Simon Willison 2026-11-12 06:03 0 阅读 机器翻译

2026年11月11日 - Link Blog

Gemini Decisions API(via)Google 凭借“Gemini Decision 1”加入 Jev 和 OpenAI 所在的决策模型类别。这个新模型 API 与 Jev 几乎完全相同:它支持同样三种类型的问题(是/否、选择、评分量规打分),甚至沿用了 Jev 指代是/否问题的术语“Noul”。

它的定价与 GPT-6 Luna 决策模型相同:每百万输入 Token 收费 10 美分,输出免费。Jev 仍是三者中最便宜的,为每百万 4.2 美分。

Jev 只支持文本(以及 JSON)。OpenAI 可以处理文本和图像。Gemini Decision 1 则是完全多模态:覆盖文本、图像、音频、视频和 PDF。

我用视频对它做了测试,把下面这些问题连同我的 Yonkers 踩踏机器人视频文件一起传给了它:

dragon:这段视频里有龙吗?robot:这段视频里有机器人吗?france:这段视频的场景在法国吗?usa:这段视频的场景在美国吗?

对于是/否问题,Jev 和 OpenAI 都会返回 1.0 或 0.0,但 Gemini 坚持要显得没那么有把握:

{ "dragon": { "type": "yes_no", "probability": 0.04976021841520312 }, "robot": { "type": "yes_no", "probability": 0.9234055387792485 }, "france": { "type": "yes_no", "probability": 0.08329171353166241 }, "usa": { "type": "yes_no", "probability": 0.9230459995157706 } }

发布于 2026 年 11 月 11 日晚上 10:03

这是 Simon Willison 发布的一篇链接帖,发布于 2026 年 11 月 11 日。

每月赞助我 10 美元,即可获得一份精选邮件摘要,内容涵盖当月最重要的 LLM 进展。

付钱给我,让我少给你发点东西!

译文由英文原文机器翻译,术语按行业惯例统一,可在文末展开原文核对。

查看英文原文(Gemini Decisions API)

11th November 2026 - Link Blog

Gemini Decisions API (via) Google join Jev and OpenAI in the decision model category with "Gemini Decision 1". This new model API is almost identical to Jev: it supports the same three types of questions (yes/no, choice, and rubric-scoring) and even reuses Jev's term "Noul" for a yes/no question.

It's priced the same as the GPT-6 Luna decision model: 10 cents per million input tokens, and output is free. Jev remains the cheapest of the three, at 4.2 cents per million.

Jev is text (and JSON) only. OpenAI can handle text and images. Gemini Decision 1 is fully multi-model: it covers text, images, audio, video, and PDFs.

I tested it against video by passing these questions to it along with my Yonkers stomping robot video file:

dragon: Does this video have a dragon in it? robot: Does this video have a robot in it? france: Is this video set in France? usa: Is this video set in the USA?

Jev and OpenAI both return 1.0 or 0.0 for yes/no questions, but Gemini insists on being a little less confident than that:

{ "dragon": { "type": "yes_no", "probability": 0.04976021841520312 }, "robot": { "type": "yes_no", "probability": 0.9234055387792485 }, "france": { "type": "yes_no", "probability": 0.08329171353166241 }, "usa": { "type": "yes_no", "probability": 0.9230459995157706 } }

Posted 11th November 2026 at 10:03 pm

This is a link post by Simon Willison, posted on 11th November 2026.

Sponsor me for $10/month and get a curated email digest of the month's most important LLM developments.

Pay me to send you less!

原文为英文,正文由自动抓取并保留原文。

本文由采集器自动同步

原文发布于 Simon Willison(Gemini Decisions API),标题、摘要与正文由机器翻译,术语按行业惯例统一,未经人工逐句校订。

阅读原文 ↗