HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks
Home/Orgs/allenai/allenai/Olmo-3-32B-Think-DPO

allenai/Olmo-3-32B-Think-DPO

family olmo · 2 variants · unknown

provider
allenai
released
—
huggingface
https://huggingface.co/allenai/Olmo-3-32B-Think-DPO
best scores
MATH: 96.2AIME 2024: 80.6AIME 2025: 78.1BBH: 88.8AGIEval: 89.2HumanEval: 91.5MBPP: 68.3LiveCodeBench: 83.3IFEval: 93.8IFBench: 68.1MMLU: 86.4PopQA: 33.1GPQA: 57.5AlpacaEval: 69.1

Best scores per benchmark

benchmarkcategorybest score
MATHmath96.2
IFEvalinstruction-following93.8
HumanEvalcoding91.5
AGIEvalreasoning89.2
BBHreasoning88.8
MMLUknowledge86.4
LiveCodeBenchcoding83.3
AIME 2024math80.6
AIME 2025math78.1
AlpacaEvalchat69.1
MBPPcoding68.3
IFBenchinstruction-following68.1
GPQAreasoning57.5
PopQAknowledge33.1

Variants (2)

variantkindquantbitsvRAMdlorgyear
Olmo-3-32B-Thinkoriginalbf1616.011,352allenai2026
Olmo-3.1-32B-Thinkoriginalbf1616.06,444allenai2026
Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.