LesenForschungRadarAnlageframework
Anmelden / Registrieren
← Zurück zur Lektüre
NewsThema · Mercor会计基准Wesentlich1 Min. LesezeitDisconfirmAI Redaktion

AI now beats licensed accountants on bookkeeping tasks, but not on closing the books

AI tops CPAs on simplified accounting tasks, yet no model fully solves nearly 60% of the full benchmark, so replacement stays at task level.

2026-10-02 02:59 GMT+8

On simplified tasks from the APEX Accounting Benchmark, AI now solves them almost flawlessly, surpassing the roughly 37% average of 12 licensed CPAs with 5.5 years of experience — a comparison drawn from a Mercor study.

Eighteen months earlier, the best models still scored below that human level, leaving AI's replacement ability under the accountants'.

On the full APEX benchmark, which spans 160 tasks across 10 simulated companies, the leading model, Claude Opus 5.5, meets 61.8% of grading criteria, followed by Fable 5.1 at 61.0% and GPT-6 Astra at 57.9%. Mercor says no model fully solved almost 60% of the tasks.

Mercor acknowledges the tasks test exactly what AI does best, hunting down details and following instructions precisely, and leave out client conversations, colleague coordination and years of accumulated business context.

Quellen

  • mercor.com↗
Thema

Mercor会计基准

【背景】Mercor发布的APEX基准测试旨在量化AI在知识工作中的实际生产力。最新数据显示,在简化的结构化会计任务中,Claude Opus 5等模型以100%准确率和低于10分钟的耗时,大幅超越平均得分37%的人类持证会计师,单次评分标…

Anmelden / Registrieren
LesenForschungRadarAnlageframework

5 Schritte1 BeiträgeAktualisiert 2026-10-05 13:31

Wie es dazu kam

  1. 2025-10-01TIME报道Mercor发布APEX指数,GPT-5在综合知识工作中得分为64.2%,但仅两项任务获满分,指出其难以处理开放式问题。time.com ↗
  2. 2026-02-09AOL报道Mercor CEO Brendan Foody称,尽管AI代理在咨询任务首次尝试成功率不足25%,但进步迅速,预计年底可达50%。aol.com ↗
  3. 2026-10-02The Decoder披露完整APEX会计基准结果:Claude Opus 5.5领先(61.8%),但无模型能独立解决近60%的任务,强调仍需监督。the-decoder.com ↗

Was noch ungeklärt ist

  • 当模型在完整基准中的得分突破多少阈值时,企业会开始实质性削减初级会计岗位?
  • 如何量化“客户沟通”和“专业判断”等非结构化工作在AI辅助下的效率提升幅度?
  • 监管层面是否会针对AI生成的财务数据建立新的审计责任归属标准?
Vollständiges Thema lesen →
Recherche fortsetzen →