What does Artificial Analysis · Humanity's Last Exam measure?

Artificial Analysis · Humanity's Last Exam is a knowledge benchmark in the BenchGecko catalog. 66 AI models have been tested on it. Scores range from 1.1 to 61.4 out of 100.

Which model leads on Artificial Analysis · Humanity's Last Exam?

Claude Opus 5.5 from Anthropic leads Artificial Analysis · Humanity's Last Exam with a score of 61.4. The median score across 66 tested models is 20.0.

Is Artificial Analysis · Humanity's Last Exam saturated?

No · the top score is 61.4 out of 100 (61%). There is still meaningful room for improvement on Artificial Analysis · Humanity's Last Exam.

Does Artificial Analysis · Humanity's Last Exam predict performance on other benchmarks?

Yes · Artificial Analysis · Humanity's Last Exam scores correlate 0.99 with Terminal Bench across 6 shared models. Models that do well on Artificial Analysis · Humanity's Last Exam tend to do well on Terminal Bench.

How often is Artificial Analysis · Humanity's Last Exam data refreshed?

BenchGecko pulls updates daily. New model scores on Artificial Analysis · Humanity's Last Exam appear as soon as they are published by Epoch AI or the model provider.

Benchmark · KnowledgeCompetitive

Artificial Analysis · Humanity's Last Exam

Name: Artificial Analysis · Humanity's Last Exam Benchmark
Creator: BenchGecko
License: https://creativecommons.org/licenses/by/4.0/

Updated 2026-09-29

Models tested

Top score

61.4

Claude Opus 5.5

Median

20.0

min 1.1

Top-5 spread

σ 3.1

Competitive

The Frontier

Best score over time · one chart, every benchmark

Chart type

Frontier on Artificial Analysis · Humanity's Last Exam rose from 3.8 to 61.4 in 21 months · +57.6 points · latest leader Claude Opus 5.5 from Anthropic.

Pink dots = frontier records · 7 totalClick to open model page

Full rankings

66 models tested · sorted by score

#	Model	Score	Price
1	Claude Opus 5.5· Anthropic	61.4	$4.00
2	Claude Fable 5.1· Anthropic	59.1	$10.00
3	Claude Sonnet 5.5· Anthropic	55.0	$2.00
4	GPT-6 Astra· OpenAI	54.7	$10.00
5	GPT-6.1 Sol· OpenAI	52.9	$2.00
6	MiMo-V2.6-Pro· xiaomi	49.4	$0.43
7	Gemini 3.8 Flash· Google DeepMind	47.8	$0.75
8	Gemini 3.1 Pro Preview· Google DeepMind	47.0	$2.00
9	Kimi K3· moonshotai	46.9	$2.70
10	Grok 4.7· xAI	43.1	$2.00
11	Qwen3.8 Max (0902)· Alibaba Qwen	43.1	$2.00
12	GPT-5.6 Terra· OpenAI	42.9	$2.00
13	GPT-5.3-Codex· OpenAI	42.5	$1.75
14	Qwen3.8 2.4T A95B· Alibaba Qwen	42.4	$2.00
15	GLM 5.3· z-ai	42.3	$1.40
16	DeepSeek V4 Pro· DeepSeek	41.0	$0.21
17	GLM 5.3 Flash· z-ai	39.9	$0.15
18	DeepSeek V4.1 Flash· DeepSeek	39.2	$0.15
19	MiniMax M3· minimax	39.0	$0.30
20	GPT-6 Luna· OpenAI	38.5	$0.10
21	Qwen3.7 Plus· Alibaba Qwen	35.6	$0.32
22	MiMo-V2.6-Flash· xiaomi	35.1	$0.14
23	Kimi K2.7 Code· moonshotai	35.0	$0.61
24	Qwen3.8 27B· Alibaba Qwen	33.9	$0.42
25	LongCat 2.0· meituan	33.7	$0.30
26	Hy3· tencent	33.5	$0.13
27	Solar Pro 4· upstage	29.2	$0.09
28	Nemotron 3 Ultra· NVIDIA	28.4	$0.50
29	Solar Mini 4· upstage	25.8	$0.05
30	Gemma 4 31B· Google DeepMind	23.6	$0.09
31	Qwen3.6 35B A3B· Alibaba Qwen	22.2	$0.15
32	Nemotron 3 Super· NVIDIA	20.8	$0.08
33	o3· OpenAI	20.1	$2.00
34	Qwen3.5 397B A17B· Alibaba Qwen	19.8	$0.55
35	gpt-oss-120b· OpenAI	19.6	$0.04
36	Gemma 4 26B A4B · Google DeepMind	19.3	$0.07
37	Gemini 3.5 Flash Lite· Google DeepMind	18.8	$0.30
38	Qwen3.5-122B-A10B· Alibaba Qwen	15.9	$0.26
39	Trinity Large Thinking· arcee-ai	15.8	$0.25
40	Mistral Medium 3.5· Mistral AI	13.8	$1.50
41	Qwen3 Next 80B A3B Instruct· Alibaba Qwen	12.6	$0.09
42	Command A+· Cohere	12.0	$0.30
43	Mercury 2.5· inception	11.8	$0.04
44	Granite 4.2 8B· ibm-granite	11.2	$0.06
45	North Mini Code (free)· Cohere	11.1	$0.00
46	gpt-oss-20b· OpenAI	11.0	$0.02
47	N Nanbeige4.1 3B· Nanbeige	10.9	—
48	DeepSeek V4 Pro 0813· DeepSeek	10.6	$0.66
49	Nemotron 3.5 Lightning· NVIDIA	10.6	$0.06
50	Qwen3 Coder Next· Alibaba Qwen	10.1	$0.12
51	Mistral Small 4· Mistral AI	9.9	$0.15
52	Qwen3.5 4B· Alibaba	9.9	—
53	Qwen3.5-9B· Alibaba Qwen	9.4	$0.10
54	LFM2.5-2.6B (free)· liquid	6.2	$0.00
55	Granite 4.0 Micro· ibm-granite	5.0	$0.02
56	Phi 4 Multimodal Instruct· Microsoft	5.0	—
57	Llama 4 Maverick· Meta	4.9	$0.19
58	NVIDIA Nemotron Nano 9B V2· NVIDIA	4.9	—
59	Phi 4 Mini Instruct· Microsoft	4.5	—
60	Reka Flash 3· rekaai	4.4	$0.10
61	Molmo2 8B· allenai	4.3	—
62	Command A· Cohere	4.0	$2.50
63	Llama 4 Scout· Meta	3.8	$0.10
64	Phi 4· Microsoft	3.8	$0.07
65	Qwen3.5 2B· Alibaba	2.6	—
66	Qwen3.5 0.8B· Alibaba	1.1	—