Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Coding on LiveCodeBench

79Task Accuracy

GPT-OSS-20B (high)

9.52827.56445.663.636Feb 10, 2026
Updated 4d ago

Evaluation Results

MethodLinks
2026.02
790.690.640.62
2026.02
770.710.660.64
2026.02
770.670.640.6
2026.02
35.1---
2026.02
300.810.830.59
2026.02
28.6---
2026.02
26.8---
2026.02
23---
2026.02
19.7---
2026.02
19.3---
2026.02
18.9---
2026.02
18.5---
2026.02
18.3---
2026.02
18.1---
2026.02
17.8---
2026.02
17.8---
2026.02
17.1---
2026.02
16.8---
2026.02
16.1---
2026.02
15.2---
2026.02
150.90.840.61
2026.02
140.910.860.64
2026.02
12.2---