ClawArena-Team - Workspace-Permission Precision: leaderboard
ClawArena-Team (2026) measures how well one language model manages a team of subagents: over 41 multi-turn multimodal scenarios, a text-only main agent must create, empower and schedule a fixed local pool of text, vision and audio subagents, with execution-based checks and no LLM judge. The board is workspace-permission precision, the files each subagent accesses divided by the files it was granted, in percent; higher is better.
Metric: Workspace-permission precision (%): per subagent, files accessed divided by files granted, averaged; 41 multi-turn multimodal scenarios with 258 evaluation rounds; a text-only main agent with partial workspace access creates, empowers and schedules a fixed locally served pool of LLM, VLM and omni subagents; execution-based checks, no LLM judge; higher is better. Source: arxiv.org. Saturation forecast: Around June 2027. 12 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 | 49.2 | #27 |
| 2 | GPT-5.5 | 47.5 | #49 |
| 3 | DeepSeek V4 Pro | 46.3 | #188 |
| 4 | Gemini 3.5 Flash | 45.7 | #101 |
| 5 | GLM-5.2 | 44.4 | #138 |
| 6 | Claude Sonnet 4.6 | 43.2 | #161 |
| 7 | Kimi K2.6 | 41.9 | #178 |
| 8 | GPT-5.4 | 40.8 | #123 |
| 9 | Qwen 3.6 27B | 40.7 | #341 |
| 10 | Gemma 4 31B | 37.9 | #394 |
| 11 | Gemini 3.1 Pro (Preview) | 36.6 | #94 |
| 12 | GLM-4.7 Flash | 11.7 | #1259 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=clawarena-team-workspace-permission-precision · How It Works · Data refreshed daily, snapshot 2026-11-13.