読書リサーチレーダー投資フレームワーク
ログイン / 新規登録
← 読解に戻る
ニューストピック · OpenAI决策接口重大1 分で読めるDisconfirmAI 編集部

OpenAI Launches Decisions API, Cuts Tiers

OpenAI released the Decisions API for fast text/image classification and reduced its paid API tiers from five to three.

2026-10-07 23:36 GMT+8

OpenAI launched the Decisions API on October 7, designed for simple judgment tasks involving text and images.

Currently in public beta and supporting only the gpt-6-luna model, the company claims it runs about ten times faster than the Responses API. Pricing is set at $0.10 per million input tokens with free output. It returns yes/no probabilities, category picks, or scale ratings, targeting use cases like damage detection and customer inquiry routing.

Simultaneously, OpenAI simplified its paid API structure, cutting tiers from five to three: Build, Launch, and Grow. Organizations automatically upgrade when total credit purchases hit the next threshold, with monthly usage limits of $500, $5,000, and $200,000 respectively.

ソース

  • the-decoder.com↗
トピック

OpenAI决策接口

OpenAI推出Decisions API,旨在通过GPT-6 Luna模型提供毫秒级、低成本的分类与决策服务,直接对标TypeSafe的Jev模型。该接口支持文本和图像输入,将响应时间从秒级压缩至百毫秒级,并简化了计费模式。尽管在部分基准…

3 ステップ

ログイン / 新規登録
読書リサーチレーダー投資フレームワーク
5 件
更新日 2026-10-25 19:41

経緯

  1. 2026-09-30OpenAI CEO Sam Altman在DevDay上首次透露Decisions API,称其功能类似TypeSafe的Jev,用于快速低成本分类。techcrunch.com ↗
  2. 2026-10-06OpenAI发布Decisions API公测版,基于gpt-6-luna模型,支持图像输入,按输入token计费。simonwillison.net ↗
  3. 2026-10-22OpenAI正式向所有客户开放Decisions API,推理延迟降至100-180毫秒,并新增GPT-Live原生集成等功能。openai.com ↗

未解決の問題

  • 在AIM System One Index等公开基准测试中,OpenAI Decision Model排名低于Liquid的D1 Ultra和Base Lab的Elephant-1,其实际生产环境中的校准度如何?
  • TypeSafe CEO Diogo Almeida强调合成数据是护城河,OpenAI依靠Luna模型的通用能力能否在特定决策任务上长期保持成本效益优势?
  • 随着延迟进一步降低至60毫秒以下,是否会催生如视频帧级实时决策等新用例,从而改变现有智能体架构?

このスレッド内の他の記事

  • OpenAI Fully Opens Decisions API: Millisecond-Level Decision-Making Based on GPT-6 Luna Targets Fast, Low-Cost AI Market2026-10-23
  • OpenAI Makes Decisions API Widely Available, Cutting Latency to Sub-Second Levels2026-10-23
トピック全文を読む →
リサーチを続ける →