CoCoBench - Task Allocation: leaderboard
Metric: Task success rate (%; task-allocation instances, dividing independent subtasks without duplicated work; 897 oracle-validated executable household multi-agent task instances with two to four agents, centralized planner with image observations, task-specific step budgets). Source: arxiv.org. Saturation forecast: Around December 2026. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 VL 8B Instruct | 95 |
| 2 | Qwen 3.5 9B | 91.3 |
| 3 | Claude Haiku 4.5 | 89.4 |
| 4 | GPT-5.4 Mini | 81.7 |
| 5 | Qwen 3.6 Plus | 78.4 |
| 6 | Kimi K2.6 | 71.6 |
| 7 | GPT-5.6 Sol | 57.3 |
| 8 | Claude Opus 4.8 | 38.1 |
Interactive version: theaggregate.ai/benchmark?slug=cocobench-task-allocation · How It Works · Data refreshed daily, snapshot 2026-09-26.