What does MUSR measure?

MUSR is a knowledge benchmark in the BenchGecko catalog. 78 AI models have been tested on it. Scores range from 1.4 to 28.7 out of 100.

Which model leads on MUSR?

DeepSeek R1 Distill Qwen 14B from DeepSeek leads MUSR with a score of 28.7. The median score across 78 tested models is 9.6.

No · the top score is 28.7 out of 100 (29%). There is still meaningful room for improvement on MUSR.

Does MUSR predict performance on other benchmarks?

Yes · MUSR scores correlate 0.91 with Lmca across 6 shared models. Models that do well on MUSR tend to do well on Lmca.

How often is MUSR data refreshed?

BenchGecko pulls updates daily. New model scores on MUSR appear as soon as they are published by Epoch AI or the model provider.

Benchmark · KnowledgeCompetitive

MUSR

Name: MUSR Benchmark
Creator: BenchGecko
License: https://creativecommons.org/licenses/by/4.0/

Updated 2025-10-17

Models tested

Top score

28.7

DeepSeek R1 Distill Qwen 14B

Median

9.6

min 1.4

Top-5 spread

σ 3.7

Competitive

The Frontier

Best score over time · one chart, every benchmark

Chart type

Only 21 models have been tested on MUSR · not enough history to compute a frontier yet.

Pink dots = frontier records · 1 totalClick to open model page

Full rankings

78 models tested · sorted by score

#	Model	Score	Price
1	DeepSeek R1 Distill Qwen 14B· DeepSeek	28.7	n/a
2	Hermes 3 70B Instruct· nousresearch	23.4	$0.70
3	Llama 3 8B Instruct· Meta	19.9	$0.14
4	Qwen2-72B· Alibaba Qwen	19.7	n/a
5	U Stable Beluga 2· Unknown	18.6	n/a
6	Llama 3.1 70B Instruct· Meta	17.7	$0.40
7	D Dolphin 2.9.1 Yi 1.5 34b· DPHN	17.0	n/a
8	DeepSeek R1 Distill Qwen 32B· DeepSeek	16.1	n/a
9	R1 Distill Qwen 32B· DeepSeek	16.1	$0.29
10	Meta Llama 3 8B· Meta	16.0	n/a
11	Llama 3.3 70B Instruct· Meta	15.6	$0.10
12	WizardLM-2 8x22B· Microsoft	14.5	$0.62
13	GLM 4 32B · z-ai	14.2	$0.10
14	Phi 2· Microsoft	13.8	n/a
15	Qwen2.5 Coder 32B Instruct· Alibaba Qwen	13.7	$0.66
16	Qwen2 VL 7B Instruct· Alibaba	13.6	n/a
17	Qwen2.5 32B Instruct· Alibaba	13.5	n/a
18	Magnum v4 72B· anthracite-org	13.4	$3.00
19	DeepSeek R1 Distill Llama 8B· DeepSeek	13.3	n/a
20	R1 Distill Llama 70B· DeepSeek	13.3	$0.80
21	Phi 3 Mini 4k Instruct· Microsoft	13.1	n/a
22	Gpt2· OpenAI	13.0	n/a
23	HA Qwen2.5 72B Instruct Abliterated· HuiHui AI	12.3	n/a
24	Qwen2 1.5B Instruct· Alibaba	12.0	n/a
25	Qwen2.5 72B Instruct· Alibaba Qwen	11.7	$0.36
26	Phi 4· Microsoft	11.4	$0.07
27	L Vicuna 7b V1.5· LMSYS	11.4	n/a
28	Gemma 2 2b· Google DeepMind	11.3	n/a
29	Hermes 2 Pro - Llama-3 8B· nousresearch	11.3	$0.14
30	D Distilgpt2· DistilBERT	11.2	n/a
31	QwQ 32B· Alibaba Qwen	11.1	$0.15
32	Gemma 2B· Google DeepMind	11.0	n/a
33	Mistral 7B V0.1· Mistral AI	10.7	n/a
34	Pythia 160m· eleutherai	10.7	n/a
35	Qwen2.5 14B Instruct· Alibaba	10.6	n/a
36	Phi 3.5 Mini Instruct· Microsoft	10.1	n/a
37	HF SmolLM2 135M· Hugging Face TB	10.0	n/a
38	Gemma 2 9B· Google DeepMind	9.7	$0.03
39	U Yi 6B· Unknown	9.7	n/a
40	Qwen2.5 Coder 7B Instruct· Alibaba Qwen	9.5	$0.03
41	Qwen2.5 Coder 7B Instruct· Alibaba	9.5	n/a
42	Gemma 2 27B· Google DeepMind	9.1	$0.65
43	Llama 3.1 8B Instruct· Meta	8.5	$0.02
44	Qwen2.5 7B Instruct· Alibaba Qwen	8.4	$0.10
45	Mistral 7B Instruct V0.2· Mistral AI	7.6	n/a
46	Qwen2.5 3B Instruct· Alibaba	7.6	n/a
47	Falcon-180B· TII	7.5	n/a
48	Qwen2 7B Instruct· Alibaba	7.4	n/a
49	Llama 3-70B· Meta	7.1	n/a
50	Gemma 2 2b It· Google DeepMind	7.1	n/a
51	Qwen2.5 Coder 14B Instruct· Alibaba	7.0	n/a
52	Phi 4 Mini Instruct· Microsoft	6.5	$0.08
53	Phi 4 Mini Instruct· Microsoft	6.5	n/a
54	Gpt2 Medium· OpenAI	6.2	n/a
55	Mistral 7B Instruct v0.1· Mistral AI	6.1	$0.11
56	Gpt2 Large· OpenAI	5.7	n/a
57	Qwen2 0.5B· Alibaba	4.6	n/a
58	T TinyLlama 1.1B Chat V1.0· TinyLlama	4.3	n/a
59	U INTELLECT-1· Unknown	4.1	n/a
60	Llama 3.2 3B Instruct (free)· Meta	3.8	$0.00
61	Llama 2 7b Hf· Meta	3.8	n/a
62	HF SmolLM2 135M Instruct· Hugging Face TB	3.7	n/a
63	DeepSeek R1 Distill Qwen 7B· DeepSeek	3.5	n/a
64	Phi-1.5· Microsoft	3.4	n/a
65	Llama 2 7b Chat Hf· Meta	3.3	n/a
66	Qwen2.5 1.5B Instruct· Alibaba	3.2	n/a
67	U RedPajama-INCITE-7B-Base· Unknown	3.0	n/a
68	DeepSeek R1 Distill Qwen 1.5B· DeepSeek	3.0	n/a
69	U StarCoder 2 15B· Unknown	2.9	n/a
70	U MPT-30B· Unknown	2.9	n/a
71	Gpt Neo 125m· eleutherai	2.6	n/a
72	Qwen2 0.5B Instruct· Alibaba	2.4	n/a
73	Llama 3.1 405B· Meta	2.2	n/a
74	LLaMA-13B· Meta	2.0	n/a
75	Llama 3.2 1B Instruct· Meta	1.9	$0.03
76	Meta Llama 3 8B Instruct· Meta	1.6	n/a
77	Llama 3.2 3B Instruct· Meta	1.4	$0.05
78	Qwen2.5 0.5B Instruct· Alibaba	1.4	n/a