Expert Consulting Deep Research - Verifier Pass Rate: leaderboard
Deep research agents on 70 management-consulting prompts written by former MBB and Big Four consultants, each with 2-4 input files and embedded traps such as inconsistent units or a footnote that changes the answer (Asthana et al., 2026). Share of the task-specific deterministic verifiers (about 15 per task) that the deliverable passes, in percent; higher is better.
Metric: Mean share of task-specific deterministic binary verifiers passed (%; about 15 per task; 70 SME-authored management-consulting prompts with 2-4 input files and embedded cognitive traps; one attempt per prompt). Source: arxiv.org. Saturation forecast: Around February 2027. 3 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 48.6 | #100 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=expert-consulting-deep-research-verifier-pass-rate · How It Works · Data refreshed daily, snapshot 2026-11-12.