SUPER — leaderboard

Metric: Expert Accuracy (%). Source: huggingface.co. 9 models tracked.

Top models

#ModelScore
1SWE-Agent + gpt-4o-2024-08-0616.3
2React-Super + gpt-4o-2024-08-0614.4
3React + gpt-4o-2024-08-0612.2
4React-Super + Llama 3.1 70B6.1
5React-Super + gpt-4o-mini-2024-07-185.6
6SWE-Agent + Llama 3.1 70B5.6
7SWE-Agent + gpt-4o-mini-2024-07-183.3
8React-Super + Mixtral-8x22B-Instruct-v0.13.3
9SWE-Agent + Mixtral-8x22B-Instruct-v0.11.1

Interactive version: theaggregate.ai/benchmark?slug=super · How the rankings work · Data refreshed daily, snapshot 2026-07-22.