CodeTransOcean - LLMTrans — leaderboard

CodeTransOcean sub-benchmark evaluating LLM-specific code translation capabilities. Evaluated by Debug Success Rate (DSR).

Metric: DSR (%). Source: yuchen814.github.io. 4 models tracked.

Top models

#ModelScore
1GPT-3.5 (3rd Debug)52.57
2GPT-3.5 (One-shot)50.29
3GPT-3.5 (Zero-shot)48.57
4GPT-3.5 (CoT)48.29

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