Atlas

Benchmarks

← All benchmarks

Code Migration - COBOL

Code · 2026-06-16

The COBOL split of Code Migration. This child benchmark separates a source-reported subtask or subtrack from the parent aggregate so scores at different grains do not share one benchmark_slug.

Top models (higher is better)

ModelScore
Claude Haiku 5.570.0
Opus 4.770.0
Claude Opus 570.0
Claude Sonnet 5.570.0
DeepSeek V4.1 Flash 101770.0
DeepSeek V4.1 Flash70.0
DeepSeek V4 Pro 081370.0
Gemini 4 Argon70.0
GLM-5.270.0
GLM-5.370.0
GPT-5.570.0
GPT-5.6 Luna70.0
GPT-5.6 Sol70.0
GPT-5.6 Terra70.0
GPT-6.1 Sol70.0
GPT-6 Sol70.0
Grok 4.570.0
Grok 4.770.0
Hy4 Preview70.0
MiMo-V2.6-Flash70.0
MiMo-V2.6-Pro70.0
Mistral Large 4 Preview70.0
Muse Wonder Full70.0
Muse Wonder70.0
Opus 4.868.6
Sonnet 4.668.6
Sonnet 568.6
Gemini 3.7 Flash68.6
Gemini 3.8 Flash68.6
GPT-5.468.6
GPT-6 Astra68.6
Muse Spark 1.168.6
Muse Spark 1.268.6
Muse Spark 1.368.6
Gemini 3.6 Flash65.7
Kimi K2.7 Code61.3
Ember-160.0
GLM-5.3 Flash60.0
Grok 4.660.0
Kimi K360.0
Loading Atlas data…