granite-4.0-h-tiny-base: benchmark results

Provider: IBM. Access: Open.

Aggregate Rating 1406 ± 34, rank #5994 of 10552 rated models, from 32 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Lara Mini Base Report (base-model protocol) - MBPP (3-shot)58.4pass@1 (execution) (%)90.7
Lara Mini Base Report (base-model protocol) - CRUXEval-O (2-shot)40pass@1, direct prediction, official answer parser (%)87
Lara Mini Base Report (base-model protocol) - Global MMLU, All Languages (5-shot)52.9Accuracy (%) averaged over all languages/variants of the ben86.1
Lara Mini Base Report (base-model protocol) - MBPP+ (3-shot)58.5pass@1; must pass MBPP's own and the EvalPlus tests (%)84.3
Lara Mini Base Report (base-model protocol) - ARC-Challenge (25-shot)61.7Accuracy normalized by answer length (%), English benchmark83.3
Lara Mini Base Report (base-model protocol) - HellaSwag (10-shot)77.7Accuracy normalized by answer length (%), English benchmark81.5
Lara Mini Base Report (base-model protocol) - MMLU (5-shot)65.5Accuracy (%), English benchmark80.6
Lara Mini Base Report (base-model protocol) - PIQA (0-shot)79.5Accuracy normalized by answer length (%), English benchmark79.6
Lara Mini Base Report (base-model protocol) - T-HellaSwag, All Languages (0-shot)43.9Accuracy normalized by answer length (%) averaged over all l79.6
Lara Mini Base Report (base-model protocol) - WinoGrande (5-shot)73.5Accuracy (%), English benchmark79.6
Lara Mini Base Report (base-model protocol) - CRUXEval-I (2-shot)36.6pass@1, direct prediction, official answer parser (%)77.8
Lara Mini Base Report (base-model protocol) - Global PIQA (non-parallel), All Languages (0-shot)58Accuracy (%) averaged over all languages/variants of the ben77.8

Interactive version: theaggregate.ai/model?slug=granite-4-0-h-tiny-base · How It Works · Data refreshed daily, snapshot 2026-11-16.