What does Artificial Analysis · SciCode measure?

Artificial Analysis · SciCode is a knowledge benchmark in the BenchGecko catalog. 55 AI models have been tested on it. Scores range from 14.4 to 66.9 out of 100.

Which model leads on Artificial Analysis · SciCode?

Claude Opus 5.5 from Anthropic leads Artificial Analysis · SciCode with a score of 66.9. The median score across 55 tested models is 46.6.

Is Artificial Analysis · SciCode saturated?

No · the top score is 66.9 out of 100 (67%). There is still meaningful room for improvement on Artificial Analysis · SciCode.

Does Artificial Analysis · SciCode predict performance on other benchmarks?

Yes · Artificial Analysis · SciCode scores correlate 0.93 with LiveBench · Reasoning across 7 shared models. Models that do well on Artificial Analysis · SciCode tend to do well on LiveBench · Reasoning.

How often is Artificial Analysis · SciCode data refreshed?

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

Benchmark · KnowledgeSettled

Artificial Analysis · SciCode

Name: Artificial Analysis · SciCode Benchmark
Creator: BenchGecko
License: https://creativecommons.org/licenses/by/4.0/

Updated 2026-10-13

Models tested

Top score

66.9

Claude Opus 5.5

Median

46.6

min 14.4

Top-5 spread

σ 2.2

Competitive

The Frontier

Best score over time · one chart, every benchmark

Chart type

Frontier on Artificial Analysis · SciCode rose from 31.7 to 66.9 in 18 months · +35.2 points · latest leader Claude Opus 5.5 from Anthropic.

Pink dots = frontier records · 8 totalClick to open model page

Full rankings

55 models tested · sorted by score

#	Model	Score	Price
1	Claude Opus 5.5· Anthropic	66.9	$4.00
2	Claude Fable 5.1· Anthropic	63.1	$10.00
3	Gemini 4 Argon· Google DeepMind	61.8	$2.00
4	Claude Sonnet 5.5· Anthropic	61.0	$2.00
5	MiMo-V2.6-Pro· xiaomi	60.9	$0.43
6	Kimi K3· moonshotai	59.5	$0.45
7	GLM 5.3· z-ai	59.0	$0.03
8	Step 5 Preview· stepfun	58.9	$1.00
9	Gemini 3.1 Pro Preview· Google DeepMind	58.7	$2.00
10	Grok 4.7· xAI	57.4	$2.00
11	Gemini 3.8 Flash· Google DeepMind	56.6	$0.75
12	GPT-6 Astra· OpenAI	56.5	$10.00
13	Claude Haiku 5.5· Anthropic	55.0	$0.10
14	GPT-5.6 Terra· OpenAI	55.0	$2.00
15	GPT-6 Luna· OpenAI	54.6	$0.10
16	GPT-6.1 Sol· OpenAI	54.2	$2.00
17	Mistral Large 4· Mistral AI	54.2	$0.68
18	Qwen3.8 2.4T A95B· Alibaba Qwen	54.1	$2.00
19	Qwen3.8 Max (0902)· Alibaba Qwen	52.1	$2.00
20	DeepSeek V4.1 Flash· DeepSeek	51.9	$0.15
21	GLM 5.3 Flash· z-ai	51.6	$0.15
22	MiMo-V2.6-Flash· xiaomi	51.3	$0.14
23	DeepSeek V4 Pro· DeepSeek	51.0	$0.96
24	Hy3· tencent	48.6	$0.13
25	Kimi K2.7 Code· moonshotai	47.8	$0.67
26	Solar Mini 4· upstage	47.6	$0.05
27	MiniMax M3· minimax	47.1	$0.30
28	Qwen3.8 27B· Alibaba Qwen	46.6	$0.42
29	Qwen3.7 Plus· Alibaba Qwen	46.1	$0.32
30	Gemma 4 31B· Google DeepMind	45.5	$0.09
31	Qwen3.5 397B A17B· Alibaba Qwen	44.8	$0.55
32	Solar Pro 4· upstage	44.6	$0.09
33	Gemini 3.5 Flash Lite· Google DeepMind	41.3	$0.30
34	Trinity Large Thinking· arcee-ai	40.6	$0.25
35	Nemotron 3 Ultra· NVIDIA	40.3	$0.60
36	Mistral Medium 3.5· Mistral AI	40.2	$1.50
37	Gemma 4 26B A4B · Google DeepMind	40.0	$0.07
38	DeepSeek V4 Pro 0813· DeepSeek	39.9	$0.66
39	Qwen3.5-122B-A10B· Alibaba Qwen	39.7	$0.26
40	gpt-oss-20b· OpenAI	38.9	$0.02
41	Mistral Small 4· Mistral AI	38.8	$0.15
42	North Mini Code (free)· Cohere	38.8	$0.00
43	Command A+· Cohere	38.5	$0.30
44	Mercury 2.5· inception	38.5	$0.04
45	Granite 4.2 8B· ibm-granite	37.8	$0.06
46	Qwen3.6 35B A3B· Alibaba Qwen	36.6	$0.15
47	LongCat 2.0· meituan	36.3	$0.30
48	Nemotron 3 Super· NVIDIA	36.2	$0.08
49	Qwen3 Coder Next· Alibaba Qwen	36.2	$0.12
50	gpt-oss-120b· OpenAI	34.0	$0.04
51	Nemotron 3.5 Lightning· NVIDIA	32.1	$0.07
52	Llama 4 Maverick· Meta	31.7	$0.19
53	Qwen3.5-9B· Alibaba Qwen	29.5	$0.10
54	Llama 4 Scout· Meta	21.3	$0.10
55	LFM2.5-2.6B (free)· liquid	14.4	$0.00