Mercor GDPval Extended: leaderboard

Models complete professional-work tasks from Mercor’s held-out GDPval extension. The native mean rubric score stays separate from the original public benchmark and from all-pass task success.

Metric: Mean rubric score (%). Source: www.mercor.com. 44 models tracked.

Top models

#ModelSettingScoreOverall rank
1Gemini 4 ArgonHigh76.4#14
2Claude Opus 5.5Max75.8#7
3Claude Sonnet 5.5Max70.9#21
4Muse WonderMax70#31
5Claude Fable 5.1Max68.7#17
6Claude Haiku 5.5Max66#76
7Claude Fable 5Max65.3#28
8GPT-6 AstraMax64.1#8
9Claude Opus 5Max62.9#23
10Grok 4.6xHigh61.8#47
11Gemini 3.8 FlashHigh60.4#39
12Muse Spark OpenxHigh59.9#81
13Claude Opus 4.8Max58.9#62
14GPT-6 SolMax58.7#29
15DeepSeek V4.1 FlashMax58.6#69

No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.

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