Falcon-H1-1.5B-Base: benchmark results

Provider: TII. Released 2025-05-20. Access: Open.

Aggregate Rating 1381 ± 34, rank #6880 of 10552 rated models, from 31 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Lara Mini Base Report (base-model protocol) - XNLI (English) (0-shot)56.8Accuracy (%), English benchmark96.3
Lara Mini Base Report (base-model protocol) - MMLU-Pro (5 (CoT)-shot)40.4Exact match (%), English benchmark90.7
Lara Mini Base Report (base-model protocol) - MATH (4-shot)46.6Math-Verify equivalence, Minerva prompts, 5,000 problems ove81.5
Lara Mini Base Report (base-model protocol) - PAWS-X, All Languages (0-shot)58.4Accuracy (%) averaged over all languages/variants of the ben79.6
Lara Mini Base Report (base-model protocol) - PAWS (English) (0-shot)67.1Accuracy (%), English benchmark76.9
Lara Mini Base Report (base-model protocol) - MBPP+ (3-shot)55.3pass@1; must pass MBPP's own and the EvalPlus tests (%)75.9
Lara Mini Base Report (base-model protocol) - MBPP (3-shot)49.4pass@1 (execution) (%)68.5
Lara Mini Base Report (base-model protocol) - MMLU (5-shot)61.7Accuracy (%), English benchmark61.1
Lara Mini Base Report (base-model protocol) - CRUXEval-O (2-shot)34.1pass@1, direct prediction, official answer parser (%)58.3
Lara Mini Base Report (base-model protocol) - MGSM (8 (CoT)-shot)36.3Exact match on each language's native CoT answer phrase, ave55.6
Lara Mini Base Report (base-model protocol) - Belebele (English) (0-shot)65.9Accuracy (%), English benchmark50
Lara Mini Base Report (base-model protocol) - GSM8K (5-shot)55.6Exact match (strict extraction: first "#### N", else first \50

Interactive version: theaggregate.ai/model?slug=falcon-h1-1-5b-base · How It Works · Data refreshed daily, snapshot 2026-11-16.