MCJudgeBench - Intrinsic Inconsistency: leaderboard
Judge self-consistency on MCJudgeBench (2026), where an LLM judge sees an instruction, a candidate response and its explicit constraint list and labels each of 653 constraints from 141 multi-constraint instructions as satisfied, partly satisfied or not satisfied: the share of constraints whose verdict changes across five repeated judgments of the same instance at temperature 1.0. Percent; lower is better.
Metric: Intrinsic constraint-level inconsistency rate (%, CIR intr): share of constraints whose predicted label is not identical across 5 repeated judgments of the same instance at temperature 1.0; the judge sees an instruction, a candidate response and its explicit constraint list and labels each constraint yes, partial or no, over the 653 constraints of 141 multi-constraint instructions of MCJudgeBench; lower is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 7 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Haiku 4.5 | 2.94 | #583 | |
| 2 | Claude Sonnet 4.6 | Thinking | 5.21 | #173 |
| 3 | Gemini 3.1 Pro (Preview) | 6.58 | #99 | |
| 4 | Gemini 2.5 Flash Lite | 7.35 | #964 | |
| 5 | GPT-5.2 | Non-reasoning | 9.49 | #185 |
| 6 | Qwen 3.5 4B | Non-reasoning | 24.2 | #1241 |
| 7 | Llama 3.2 3B Instruct | 35.38 | #7241 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=mcjudgebench-intrinsic-inconsistency · How It Works · Data refreshed daily, snapshot 2026-11-21.