Vision Arena

View rankings across multimodal, generative AI models capable of understanding and processing visual inputs

Last Updated

Feb 6, 2026

Total Votes

654,886

Total Models

/

Style Control

	Rank Spread
1	1◄─►2	gemini-3-pro	1289	±9	11,297	Google	Proprietary
2	1◄─►3	gemini-3-flash	1277	±9	9,175	Google	Proprietary
3	2◄─►9	gpt-5.2-high	1257	±14	2,749	OpenAI	Proprietary
4	3◄─►8	gemini-3-flash (thinking-minimal)	1256	±10	7,313	Google	Proprietary
5	3◄─►9	gpt-5.1-high	1252	±10	7,299	OpenAI	Proprietary
6	3◄─►10	kimi-k2.5-thinking	1251	±13	2,979	Moonshot	Modified MIT
7	3◄─►10	gemini-2.5-pro	1246	±6	79,747	Google	Proprietary
8	6◄─►14	chatgpt-4o-latest-20250326	1235	±6	23,313	OpenAI	Proprietary
9	4◄─►15	gpt-5.1	1235	±9	7,974	OpenAI	Proprietary
10	3◄─►22	kimi-k2.5-instant	1231	±17	1,663	Moonshot	Modified MIT
11	8◄─►22	gemini-2.5-flash-preview-09-2025	1225	±10	5,293	Google	Proprietary
12	8◄─►22	gpt-4.5-preview-2025-02-27	1225	±11	2,925	OpenAI	Proprietary
13	8◄─►23	gpt-5.2	1223	±14	3,013	OpenAI	Proprietary
14	8◄─►22	gpt-5-chat	1222	±7	43,264	OpenAI	Proprietary
15	9◄─►25	ernie-5.0-preview-1220	1216	±11	3,623	Baidu	Proprietary
16	10◄─►23	o3-2025-04-16	1216	±7	49,181	OpenAI	Proprietary
17	10◄─►25	gemini-2.5-flash	1213	±6	48,047	Google	Proprietary
18	10◄─►25	gpt-4.1-2025-04-14	1213	±7	44,463	OpenAI	Proprietary
19	10◄─►27	qwen3-vl-235b-a22b-instruct	1211	±8	10,750	Alibaba	Apache 2.0
20	10◄─►27	gpt-5-high	1208	±8	37,581	OpenAI	Proprietary
21	10◄─►32	claude-opus-4-20250514-thinking-16k	1206	±15	1,495	Anthropic	Proprietary
22	10◄─►34	claude-sonnet-4-20250514-thinking-32k	1205	±16	1,361	Anthropic	Proprietary
23	16◄─►31	gpt-4.1-mini-2025-04-14	1201	±8	43,674	OpenAI	Proprietary
24	16◄─►32	o4-mini-2025-04-16	1199	±7	44,239	OpenAI	Proprietary
25	14◄─►37	claude-3-7-sonnet-20250219-thinking-32k	1195	±15	1,676	Anthropic	Proprietary
26	19◄─►37	o1-2024-12-17	1192	±10	3,694	OpenAI	Proprietary
27	19◄─►37	claude-opus-4-20250514	1191	±12	2,579	Anthropic	Proprietary
28	21◄─►37	gemini-2.5-flash-lite-preview-06-17-thinking	1188	±8	39,110	Google	Proprietary
29	21◄─►38	hunyuan-vision-1.5-thinking	1187	±12	2,869	Tencent	Proprietary
30	21◄─►39	qwen3-vl-235b-a22b-thinking	1186	±12	2,664	Alibaba	Apache 2.0
31	21◄─►39	claude-sonnet-4-20250514	1186	±13	2,066	Anthropic	Proprietary
32	24◄─►39	grok-4-0709	1182	±8	34,737	xAI	Proprietary
33	24◄─►39	gpt-5-mini-high	1181	±9	31,410	OpenAI	Proprietary
34	22◄─►40	qwen-vl-max-2025-08-13	1181	±12	3,454	Alibaba	Proprietary
35	25◄─►39	gemini-1.5-pro-002	1178	±8	8,902	Google	Proprietary
36	25◄─►41	claude-3-7-sonnet-20250219	1177	±9	4,674	Anthropic	Proprietary
37	25◄─►45	gemini-2.5-flash-lite-preview-09-2025-no-thinking	1173	±10	5,330	Google	Proprietary
38	29◄─►46	gemini-2.0-flash-001	1170	±7	9,875	Google	Proprietary
39	35◄─►51	gpt-4o-2024-05-13	1162	±8	23,273	OpenAI	Proprietary
40	29◄─►53	glm-4.6v	1161	±14	2,611	Z.ai	MIT
41	36◄─►51	claude-3-5-sonnet-20241022	1161	±7	10,568	Anthropic	Proprietary
42	37◄─►53	gemma-3-27b-it	1156	±8	18,534	Google	Gemma
43	38◄─►53	mistral-medium-2505	1155	±8	11,519	Mistral	Proprietary
44	37◄─►55	glm-4.5v	1154	±12	3,576	Z.ai	MIT
45	37◄─►55	step-1o-turbo-202506	1152	±14	2,037	StepFun	Proprietary
46	37◄─►56	hunyuan-large-vision	1151	±16	1,440	Tencent	Proprietary
47	39◄─►55	mistral-medium-2508	1150	±7	41,998	Mistral	Proprietary
48	39◄─►55	claude-3-5-sonnet-20240620	1146	±9	21,624	Anthropic	Proprietary
49	39◄─►55	llama-4-maverick-17b-128e-instruct	1145	±9	7,410	Meta	Llama 4
50	39◄─►58	gpt-5-nano-high	1144	±11	4,325	OpenAI	Proprietary
51	39◄─►58	step-3	1144	±12	3,558	StepFun	Apache 2.0
52	41◄─►60	mistral-small-2506	1139	±9	11,713	Mistral	Apache 2.0
53	41◄─►61	gemini-1.5-flash-002	1139	±9	7,241	Google	Proprietary
54	44◄─►63	gemini-2.0-flash-lite-preview-02-05	1133	±10	3,991	Google	Proprietary
55	44◄─►63	claude-3-5-haiku-20241022	1130	±15	1,583	Anthropic	Proprietary
56	50◄─►63	mistral-small-3.1-24b-instruct-2503	1126	±9	30,955	Mistral	Apache 2.0
57	50◄─►63	llama-4-scout-17b-16e-instruct	1125	±10	6,826	Meta	Llama
58	49◄─►63	step-1o-vision-32k-highres	1123	±12	2,833	StepFun	Proprietary
59	52◄─►63	qwen2.5-vl-72b-instruct	1121	±10	3,768	Alibaba	Qwen
60	53◄─►63	gpt-4o-2024-08-06	1118	±12	3,376	OpenAI	Proprietary
61	54◄─►64	gemini-1.5-pro-001	1117	±11	16,734	Google	Proprietary
62	52◄─►66	qwen2.5-vl-32b-instruct	1116	±15	1,490	Alibaba	Apache 2.0
63	54◄─►66	gpt-4-turbo-2024-04-09	1112	±11	13,391	OpenAI	Proprietary
64	62◄─►68	gpt-4o-mini-2024-07-18	1097	±7	17,347	OpenAI	Proprietary
65	62◄─►68	pixtral-large-2411	1093	±9	5,423	Mistral	MRL
66	61◄─►72	gpt-4.1-nano-2025-04-14	1088	±18	1,211	OpenAI	Proprietary
67	64◄─►70	qwen2-vl-72b	1085	±9	5,937	Alibaba	Qwen
68	64◄─►72	qwen-vl-max-1119	1084	±16	1,422	Alibaba	Proprietary
69	66◄─►73	gemini-1.5-flash-8b-001	1070	±10	6,243	Google	Proprietary
70	67◄─►74	claude-3-opus-20240229	1063	±10	15,565	Anthropic	Proprietary
71	66◄─►74	step-1v-32k	1063	±16	1,534	StepFun	Proprietary
72	67◄─►74	gemini-1.5-flash-001	1059	±11	13,260	Google	Proprietary
73	70◄─►78	molmo-72b-0924	1047	±13	3,048	AI2	Apache 2.0
74	69◄─►81	hunyuan-standard-vision-2024-12-31	1043	±21	809	Tencent	Proprietary
75	73◄─►81	llama-3.2-vision-90b-instruct	1032	±8	8,682	Meta	Llama 3.2
76	73◄─►82	qwen2-vl-7b-instruct	1031	±10	5,766	Aliaba	Apache 2.0
77	74◄─►83	pixtral-12b-2409	1025	±9	7,511	Mistral	Apache 2.0
78	73◄─►84	internvl2-26b	1024	±12	5,148	OpenGVLab	MIT
79	73◄─►86	amazon-nova-lite-v1.0	1020	±15	1,854	Amazon	Proprietary
80	74◄─►86	amazon-nova-pro-v1.0	1019	±13	2,335	Amazon	Proprietary
81	74◄─►85	claude-3-sonnet-20240229	1019	±11	12,314	Anthropic	Proprietary
82	77◄─►88	yi-vision	1003	±18	1,219	01 AI	Proprietary
83	78◄─►88	claude-3-haiku-20240307	1002	±12	13,380	Anthropic	Proprietary
84	76◄─►90	c4ai-aya-vision-32b	1000	±22	847	Cohere	CC-BY-NC-4.0
85	79◄─►88	molmo-7b-d-0924	996	±13	2,815	AI2	Apache 2.0
86	82◄─►88	llama-3.2-vision-11b-instruct	991	±11	4,817	Meta	Llama 3.2
87	80◄─►92	nvila-internal-15b-v1	987	±20	1,077	Nvidia	-
88	82◄─►92	llava-onevision-qwen2-72b-ov	980	±18	1,321	LLaVA	Apache 2.0
89	87◄─►92	llava-v1.6-34b	966	±12	4,531	LLaVA	Apache 2.0
90	86◄─►92	minicpm-v-2_6	964	±15	1,987	OpenBMB	Apache 2.0
91	86◄─►92	cogvlm2-llama3-chat-19b	964	±15	1,991	Zhipu AI	CogVLM2
92	87◄─►92	internvl2-4b	957	±12	3,703	OpenGVLab	MIT
93	93◄─►93	phi-3.5-vision-instruct	921	±15	2,592	Microsoft	MIT
94	94◄─►94	phi-3-vision-128k-instruct	883	±18	1,401	Microsoft	MIT

Vision Arena

Default Leaderboard Plots

Fraction of Model A Wins for All Non-tied A vs. B Battles

Confidence Intervals on Model Strength (via Bootstrapping)

Battle Count for Each Combination of Models (without Ties)

Average Win Rate Against All Other Models (Uniform Sampling and No Ties)