EvoClawBench - Baseline (nanobot): leaderboard

Metric: Execution-only mean task score (%; 100 fixture-backed tasks plus a sanity task, 502 sub-problems across coding, data, office, security, operations and document workflows, automated or hybrid graders; one run per task in the nanobot agent runtime, execution phase only; the agent solves each task directly and may not create skills). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 5 models tracked.

Top models

#ModelScore
1nanobot + GPT-5.496.13
2nanobot + MiniMax-M2.790.97
3nanobot + GPT-5.4 Mini88.71
4nanobot + DeepSeek V4 Pro77.77
5nanobot + Qwen 3.6 Plus56.45

Interactive version: theaggregate.ai/benchmark?slug=evoclawbench-baseline-nanobot · How It Works · Data refreshed daily, snapshot 2026-09-29.