Benchmark explainer
What is APEX-Agents?
Long-horizon, cross-application agent execution on professional-services tasks (investment banking, consulting, legal work).
APEX-Agents as reported in Epoch AI's Capabilities Index CSV.
How it's scored
- Metric
- normalized performance in Epoch CSV, rendered as accuracy (%)
- Score ceiling
- 100
- Construction
- 480 tasks across 33 simulated 'worlds' (160 per domain: banking/consulting/legal); designed by professionals with 5-10 yr experience at firms like McKinsey, Goldman Sachs, Latham & Watkins.
- Human baseline
- Tasks estimated ~1.8hr for a human professional; no aggregate published human Pass@1.
How trustworthy
Discrimination
100/100Does it still separate models?
Saturation headroom
98.3/100How far from ceiling / clustered at the top?
Contamination resistance
4.1/100Public vs held-out; training-leak risk.
Harness comparability
100/100Is it apples-to-apples, or mixed / vendor-optimized?
Freshness
100/100How old is the benchmark?
Contamination history: Released 2026-01; publicly available task prompts and files may overlap with web-scraped training data — flagged by authors.
Limitations: Pass@1 frontier scores <25% as of early 2026 (vs Pass@8 ~40%); closed-source models dominate; world environment fidelity limits external validity.
Who leads APEX-Agents
| Model | Score | Evidence |
|---|---|---|
| Gemini 4 Argon | 82.4 | unverified |
| Claude Sonnet 5.5 | 75.5 | unverified |
| Claude Opus 5.5 | 73.5 | unverified |
| Claude Fable 5.1 | 68.6 | unverified |
| Gemini 3.7 Flash | 67.8 | unverified |
Compare the top APEX-Agents scorers
Frequently asked questions
APEX-Agents: Long-horizon, cross-application agent execution on professional-services tasks (investment banking, consulting, legal work). 480 tasks across 33 simulated 'worlds' (160 per domain: banking/consulting/legal).
Gemini 4 Argon leads APEX-Agents at 82.4. The full leaderboard above lists every recorded measurement, not just the headline number.
49 models have recorded scores on APEX-Agents, spanning a score spread of 78.
tensor.news grades APEX-Agents A for integrity (score 85/100), ranking #31 of 65 benchmarks we assess across discrimination, saturation, contamination resistance, harness comparability, and freshness.
No — APEX-Agents still has headroom and continues to discriminate between models rather than bunching them at the ceiling.
Its test set is unknown, so contamination risk is on the table: a high score may partly reflect training-data overlap rather than capability.
Scores are reported under a consistent harness, so comparisons on APEX-Agents are reasonably apples-to-apples.
Every APEX-Agents measurement is source-backed and tagged reproduced, self-reported, or unverified; the underlying record cites arxiv.org/abs/2601.14242.