Skip to content

APEX-Agents

A professional-services agent benchmark covering long-horizon knowledge-work tasks.

8 models

Higher is better.

  1. 1Grok 4.6xAI57.5
  2. 2Apodex 1.1Apodex38.5
  3. 3Step 5 PreviewStepFun37.8
  4. 4Kimi K3Moonshot AI37.6
  5. 5Hy4 previewTencent37.1
  6. 6dots3-note PreviewDots Studio30.8
  7. 7Solar Pro 4Upstage18.7
  8. 8Solar Open 2Upstage16.6
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Agent task-completion score
Tasks
Professional-services agent tasks
Authors
Moonshot AI / APEX-Agents benchmark authors (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.