Mercor APEX-Accounting - Mean Score: leaderboard

APEX-Accounting: 100 professional month-end-close tasks across ten held-out company worlds, 1, 475 rubric criteria; a separate eleventh world is public. Native Loop-agent results. Pass@1 and mean criterion/rubric score remain separate metrics; no strongest-harness selection. Criterion-level Met/NotMet LM grading includes a scattergun penalty. The source states five-million-token/500-step limits, step retries and at most three failed-trajectory reruns. Its k=4 criterion means and separately described eight-run pass curves must not be treated as one identical repeated-trial population; native configuration remains retained. Exact native effort, API/EAP identity and score configuration remain per observation. Missing/failure and repeated-trial policies are not inferred from display sample counts. Metric: Mean rubric score (%). Source: https://www.mercor.com/apex/apex-accounting-leaderboard/

Metric: Mean rubric score (%). Source: www.mercor.com. Saturation forecast: Around October 2027. 50 models tracked. Updated 17 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 4 ArgonHigh71.5#14
2Gemini 4 CarbonHigh70.7#3
3Claude Fable 5.2Max70.4#1
4Gemini 4 HydrogenMercor Accounting run: source-id=gemini/gemini-4-hydrogen-eap, access=EAP67.6#13
5Muse WonderMax67.1#31
6Claude Fable 5.1Max66.7#17
7Claude Opus 5.5Max65.5#7
8Muse Spark 1.3Max64.9#45
9GPT-6 AstraMax64.7#8
10Claude Sonnet 5.5Max64.5#21
11GPT-6.1 SolMax64.3#12
12Claude Opus 5Max64.3#23
13GPT-6.1 TerraMax63.7#22
14Kimi K3.1 ProMax62.4#32
15Claude Haiku 5.5Max62.2#76

No result here: #2 GPT-6.2 Astra, #6 GPT-6.2 Sol.

Interactive version: theaggregate.ai/benchmark?slug=mercor-apex-accounting-mean-score · How It Works · Data refreshed daily, snapshot 2026-11-17.