Codoku Benchmark

What is Codoku?

Evaluates coding agents on solving Codoku puzzles, where typed cells in partial programs must be filled to satisfy global static and dynamic constraints.

Released

2026-09-28

Evaluates

Code & Software, General AI, Reasoning

Openness

unknown

Importer

Claire Radar

Review status

ai-reviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2609.34661

Codoku paper and code

• Paper https://arxiv.org/abs/2609.34661
• Code https://github.com/connglli/Codoku

Which sources cite Codoku?

• Codoku

Related Code & Software, General AI, Reasoning benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 2d-fet-bench
• 3DCodeBench
• 5-Dialects-BN
• A Real-World CAD Segmentation Benchmark
• A-OKVQA
• A3S-Bench
• AARRI-Bench