# OpenAI 占据本月 token 效率帕累托前沿主导地位

- 来源：Artificial Analysis (@ArtificialAnlys)
- 发布时间：2026-07-24 02:32
- AIHOT 分数：61
- AIHOT 链接：https://aihot.virxact.com/items/cmrxvey5d031yroxpn7ncmpnv
- 原文链接：https://x.com/ArtificialAnlys/status/2080360526534877537

## AI 摘要

尽管本月有 5 家以上实验室发布重要模型，OpenAI 仍占据大部分 token 效率帕累托前沿。Artificial Analysis 衡量模型在智能指数任务中产生的输出 token（含答案与推理 token），这是决定每任务成本与时间的关键因素。GPT-5.6 Sol 的多个 effort 级别主导前沿，而 Terra 和 Luna 在同等智能水平下产生更多 token。

## 正文

Despite major launches from 5+ labs this month， OpenAI occupies most of the token efficiency Pareto frontier

We measure the number of output tokens models produce per task in the Artificial Analysis Intelligence Index. Output tokens consist of answer tokens （can be thought of as how verbose the model is） and reasoning tokens （how much the model thinks before giving an answer）. Reasoning tokens in particular offer a way for models to use compute at inference time to improve responses.

Output tokens are an important determinant of both cost and time per task. Various effort levels of GPT-5.6 Sol dominate the frontier - Terra and Luna produce comparatively more tokens for any level of intelligence.