New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
BigCodeBench benchmark
AI model leaderboard for the BigCodeBench benchmark. Compare how large language models score on BigCodeBench, see the full ranking, and understand what this AI benchmark measures. Granite-4.0-H-Small currently leads with 46.23. Benchmark for general code generation — 1140 programming tasks across 139 libraries, evaluated by full unit tests.
Leaderboard
| # | Model | Organization | Score | Variant | Source |
|---|---|---|---|---|---|
| #1 | Granite-4.0-H-Small | IBM | 46.23 | pass@1 | official ↗ |
| #2 | Granite-4.0-H-Tiny | IBM | 41.06 | pass@1 | official ↗ |
| #3 | Granite-4.0-H-Micro | IBM | 37.9 | pass@1 | official ↗ |
Frequently asked questions about BigCodeBench
What is the BigCodeBench benchmark?
Benchmark for general code generation — 1140 programming tasks across 139 libraries, evaluated by full unit tests.
How is the BigCodeBench benchmark scored?
BigCodeBench is scored using the pass@1 metric, where a higher score is better. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.
Which AI model scores highest on BigCodeBench?
As of the latest reported scores on GenAIList, Granite-4.0-H-Small achieves the highest result on BigCodeBench with a score of 46.23.
Is a higher BigCodeBench score better?
Yes. On BigCodeBench a higher score indicates better performance, so models near the top of the leaderboard are the strongest.