Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

Story

Perplexity releases open-weight Decider v1.3 decision models

AI summary

Perplexity released Decider v1.3 and Decider Flash v1.3 as Apache 2.0 open-weight decision models that return probability distributions for classification tasks instead of generating text token by token. The 27B and 4B models support text and image input, offer a 256K context window, and are available through Perplexity's API; Artificial Analysis measured Decision Index v2.1 scores of 70 and 64, respectively.

In-depth insight

The release adds two independently benchmarked, openly licensed models for classification, extraction, and related decision workflows, with documented API pricing and an option for smaller local deployment.

导语:Decider v1.3:63 项分类任务准确率达 91.6%,对标 Jev 成本直降 61%。

黄仁勋「每天都在用」的 Perplexity 开始重构 Attention,把「判断」做成了一个独立模型

Decider v1.3:63 项分类任务准确率达 91.6%,对标 Jev 成本直降 61%。

    作者丨郑佳美

    编辑丨岑   峰

刚刚,黄仁勋每天都在用的「AI 导师」 Perplexity 有了新动作:正式推出 Decider v1.3,直接把 AI 系统里的判断能力单独做成了一套模型。

#Perplexity#Decider v1.3#Decider Flash v1.3#Artificial Analysis Decision Index v2.1#Qwen3.8-27B#Qwen3.5-4B#OpenRouter

Original · 9 min readRead the original
Group chat

No comments yet. Start the conversation.