IOI

International Olympiad in Informatics problems. Run by Vals AI.

Published by

Vals AI

Category

Coding

Index weight

0.5

Models

58

Data as of

9 Sept 2026

Vals AI (vals.ai).

1. 1Claude Opus 591.7%
2. 2GPT-5.6 Sol86.7%
3. 3Qwen3 873.0%
4. 4GPT-5.6 Luna72.9%
5. 5Claude Fable 572.3%
6. 6GPT-5.467.8%
7. 7GPT-5.6 Terra65.3%
8. 8GPT-5.254.8%
9. 9Muse Spark 1.249.5%
10. 10Claude Opus 4.747.1%
11. 11Qwen3 746.8%
12. 12GPT-5.3 Codex43.8%
13. 13Gemini 3 Flash39.1%
14. 14Gemini 3 Pro38.8%
15. 15DeepSeek V4 Pro35.8%

SearchAll providersAlibabaAnthropicDeepSeekGoogleMetaMiniMaxMistral AIMoonshot AIOpenAIxAIZhipu AIOpen weights58 of 58

#	Model	Score ↓	Run ↕	Index ↕	Published ↕
1	Claude Opus 5Anthropic	91.7%	–	70.0	–
2	GPT-5.6 SolOpenAI	86.7%	–	65.5	–
3	Qwen3 8Alibaba	73.0%	–	63.9	–
4	GPT-5.6 LunaOpenAI	72.9%	–	56.5	–
5	Claude Fable 5Anthropic	72.3%	–	70.4	–
6	GPT-5.4OpenAI	67.8%	–	61.3	–
7	GPT-5.6 TerraOpenAI	65.3%	–	64.2	–
8	GPT-5.2OpenAI	54.8%	–	57.3	–
9	Muse Spark 1.2Meta	49.5%	meta/muse_spark_1_2	62.6	–
10	Claude Opus 4.7Anthropic	47.1%	–	65.5	–
11	Qwen3 7Alibaba	46.8%	–	58.9	–
12	GPT-5.3 CodexOpenAI	43.8%	–	–	–
13	Gemini 3 FlashGoogle	39.1%	–	58.5	–
14	Gemini 3 ProGoogle	38.8%	–	57.0	–
15	DeepSeek V4 ProDeepSeekopen	35.8%	–	56.1	–
16	Grok 4.20xAI	30.2%	–	54.3	–
17	Grok 4xAI	26.2%	0709	57.2	–
18	Gemini 3.5 Flash LiteGoogle	26.2%	–	44.0	–
19	Claude Opus 4.5Anthropic	23.6%	–	58.0	–
20	GLM-5Zhipu AIopen	22.0%	–	57.0	–
21	GPT-5.1OpenAI	21.5%	–	53.8	–
22	GPT-5.1-Codex-MaxOpenAI	21.4%	–	–	–
23	Claude Opus 4.5Anthropic	20.3%	–	60.6	–
24	GPT-5OpenAI	20.0%	–	54.5	–
25	Claude Sonnet 4.5Anthropic	18.3%	–	52.8	–
26	Kimi K2.5Moonshot AIopen	17.7%	–	55.3	–
27	Gemini 2.5 ProGoogle	17.1%	–	54.1	–
28	Qwen3 MaxAlibaba	15.7%	–	52.7	–
29	Grok 4.3xAI	15.3%	–	50.1	–
30	GPT-5.4 nanoOpenAI	15.3%	–	48.3	–
31	Claude Opus 4.1Anthropic	12.5%	–	52.6	–
32	Grok 4 FastxAI	11.5%	–	53.0	–
33	DeepSeek V3.2DeepSeekopen	10.7%	–	50.1	–
34	GPT-5-CodexOpenAI	9.8%	–	–	–
35	Qwen3 MaxAlibaba	7.8%	–	51.8	–
36	Grok 4.1xAI	7.7%	–	47.9	–
37	GLM-4.7Zhipu AIopen	7.6%	zai/glm-4.7	54.9	–
38	GPT-5 miniOpenAI	6.8%	–	52.3	–
39	MiniMax-M2.5MiniMaxopen	6.7%	minimax/MiniMax-M2.5	53.5	–
40	Claude Sonnet 4Anthropic	6.5%	–	49.5	–
41	GPT-5.4 miniOpenAI	6.4%	–	46.0	–
42	Claude Haiku 4.5Anthropic	6.2%	–	48.0	–
43	MiniMax-M2.7MiniMaxopen	4.9%	minimax/MiniMax-M2.7	56.5	–
44	o4-miniOpenAI	4.8%	–	48.0	–
45	Claude Sonnet 4Anthropic	4.6%	–	54.0	–
46	GLM-4.6Zhipu AIopen	4.3%	zai/glm-4.6	47.6	–
47	Grok Code Fast 1xAI	4.3%	–	45.9	–
48	Mistral Large 2Mistral AIopen	4.0%	2512	41.2	–
49	Grok 4 FastxAI	3.8%	–	41.9	–
50	GPT-5.1-CodexOpenAI	3.7%	–	–	–
51	Grok 4.1xAI	3.1%	–	55.9	–
52	GLM-4.5Zhipu AIopen	2.9%	zai/glm-4.5	50.2	–
53	Gemini 2.5 FlashGoogle	2.6%	–	48.8	–
54	Labs Devstral SmallMistral AIopen	2.5%	2512	–	–
55	MiniMax-M2MiniMaxopen	2.3%	minimax/MiniMax-M2.1	55.1	–
56	Kimi K2Moonshot AIopen	1.3%	–	48.8	–
57	Magistral MediumMistral AIopen	0.7%	medium, 2509	44.1	–
58	Qwen3 235B A22BAlibabaopen	0.0%	–	48.7	–