Granite 4.2 30B.
Open weights Flagship
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| IFBench | 77.2% | #14 of 70 | Model card | |
| Arena-Hard v2.0 | 67.9% | #13 of 24 | Model card | |
| LMArena (Elo) | 1,340 | #187 of 333 | Official leaderboard | |
| MMLU-Pro | 77.6% | #90 of 151 | Model card | |
| RULER | 81.4% | #30 of 38 | Model card | |
| MMLU-ProX | 66.6% | #14 of 17 | Model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| AIME 2025 | 89.2% | #54 of 143 | Model card | |
| HMMT 2025 | 89.2% | #24 of 49 | Model card | |
| GPQA Diamond | 66.4% | #187 of 284 | Model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| LiveCodeBench v6 | 75.8% | #40 of 76 | Model card | |
| SciCode | 38.8% | #31 of 47 | Model card | |
| SWE-bench Verified | 57% | #114 of 150 | Model card | |
| Bird-SQL | 41.9% | #10 of 13 | Model card | |
| SWE-bench Multilingual | 41.9% | #68 of 70 | Model card | |
| SWE-Bench Pro | 33.3% | #77 of 77 | Model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| BFCL V4 | 61.4% | #16 of 24 | Model card | |
| tau3-bench | 62% | #10 of 12 | Model card | |
| Terminal-Bench 2.1 | 29.2% | #51 of 53 | Model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GDPval-AA v2 | 1,225 | #21 of 26 | Model card | |
| ProfBench | 42.9% | #2 of 2 | Model card |
The benchmarks the lab chose to headline, with the scores it gave.