llmrun Composite Score · Math

The best open LLMs for math, ranked on competition and research math boards such as AIME and FrontierMath. The score shown is the overall llmrun Composite Score; the highlighted column is the mathematics skill score used for the order.

45 benchmarks4 sourcesUpdated 17 Oct 2026Reference scale 2026-Q4How the Composite Score is built

Math score vs VRAM needed

This chart doesn't show proprietary models.

Open models Overall

General Coding Agents Math Science Reasoning

Models ranked by math llmrun Composite Score
# (open / overall)	Model	Coding	Agents	Math	Science	Reasoning	Benchmarks	VRAM
1 / 25	DeepSeek V4.1 Flash458 GB · 24 benchmarks	122	128	129	123	134	24	458 GB
2 / 26	GLM 5.3454 GB · 26 benchmarks	130	133	129	131	130	26	454 GB
3 / 31	DeepSeek V4 Flash 0731183 GB · 21 benchmarks	121	–	128	124	129	21	183 GB
4 / 34	GLM 5.2454 GB · 26 benchmarks	120	128	126	129	120	26	454 GB
5 / 43	GLM 5.3 Flash205 GB · 24 benchmarks	117	127	121	124	127	24	205 GB
6 / 50	Qwen3.8 27B17.9 GB · 15 benchmarks	120	123	118	112	124	15	17.9 GB
7 / 51	GLM 5.1454 GB · 14 benchmarks	115	123	117	117	116	14	454 GB
8 / 53	Inkling Small163 GB · 11 benchmarks	–	–	117	118	123	11	163 GB
9 / 56	Qwen3.6 27B17.9 GB · 11 benchmarks	102	–	114	111	105	11	17.9 GB
10 / 57	Qwen3.5 397B A17B243 GB · 8 benchmarks	–	–	112	–	117	8	243 GB
11 / 60	NVIDIA Nemotron 3 Ultra 550B A55B BF16378 GB · 17 benchmarks	104	102	112	112	116	17	378 GB
12 / 62	GLM 4.7218 GB · 13 benchmarks	101	106	111	111	111	13	218 GB
13 / 65	Qwen3.6 35B A3B22.0 GB · 11 benchmarks	105	–	109	110	104	11	22.0 GB
14 / 70	MiniMax M3258 GB · 22 benchmarks	115	112	106	118	108	22	258 GB

The range after each score is a 90% interval: given the boards a model has results on, its true score is very likely inside it. Short ranges mean many agreeing results; long ranges mean few or conflicting ones.

Why some models are missing. A model receives a score only if it has results on at least 4 benchmarks across at least 2 categories, with one being reasoning or coding, or on 3 benchmarks that narrow its 90% range to at most 15.6 points. The results must come from at least two different benchmarks: categories of one benchmark, such as LiveBench, count once. Skill scores require at least 2 benchmarks for that skill; otherwise they display as –.

VRAM is llmrun's estimate at Q4_K_M (or the smallest quantization we track) for the weights plus room for a 16K-token context. Longer contexts need more memory. Proprietary models can't run locally, so the VRAM filter hides them.

How the llmrun Composite Score is built · llmrun doesn't run these benchmarks; scores come from public sources.