CG-AL-X272

hard Codeunits & Business Logic content 62dca2f62e43…

Description

Per-model results

Models that have attempted this task
ModelAttempt 1Attempt 2Avg scoreTokens / attemptRuns
Claude Opus 5.5✓—100.0 / 10011k / 7.2k3
GPT-6 Sol✓✓90.6 / 1006.3k / 3.3k3
GPT-6.1 Sol✓✓90.6 / 1006.0k / 2.6k3
Gemini 3.8 Flash (OpenRouter)✓✓90.6 / 1007.0k / 24k3
GPT-5.6 Sol✓✓85.0 / 1006.4k / 3.8k3
Claude Sonnet 5✗✓81.3 / 10012k / 12k3
GPT-5.6 Luna✗✓81.3 / 1006.9k / 3.4k3
GPT-5.6 Terra✗✓81.3 / 1006.8k / 3.2k3
GPT-6 Luna✗✓81.3 / 1006.7k / 3.1k3
GLM 5.3 (OpenRouter)✗✓81.3 / 1006.8k / 43k2
Claude Opus 5✓✗70.0 / 10012k / 12k3
Claude Haiku 4 5✗✗62.5 / 10012k / 2.3k3
MiniMax M3✗✗52.1 / 1007.7k / 23k3