Vision Arena

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

Last Updated

Dec 16, 2025

Total Votes

573,256

Total Models

/

Style Control

	Rank Spread (Upper-Lower)
1	1◄─►2	gemini-3-pro	1309	±12	4,329	Google	Proprietary
2	1◄─►4	gemini-3-flash	1284	±25	529	Google	Proprietary
3	2◄─►7	gemini-3-flash (thinking-minimal)	1268	±26	520	Google	Proprietary
4	2◄─►8	gpt-5.1-high	1249	±14	2,197	OpenAI	Proprietary
5	3◄─►7	gemini-2.5-pro	1249	±7	65,685	Google	Proprietary
6	3◄─►10	gpt-5.1	1239	±14	2,273	OpenAI	Proprietary
7	3◄─►10	chatgpt-4o-latest-20250326	1236	±7	17,714	OpenAI	Proprietary
8	5◄─►18	gpt-4.5-preview-2025-02-27	1226	±11	2,925	OpenAI	Proprietary
9	6◄─►18	gemini-2.5-flash-preview-09-2025	1224	±10	4,979	Google	Proprietary
10	6◄─►18	gpt-5-chat	1223	±8	38,722	OpenAI	Proprietary
11	8◄─►18	o3-2025-04-16	1217	±7	45,238	OpenAI	Proprietary
12	8◄─►21	gpt-4.1-2025-04-14	1215	±7	40,356	OpenAI	Proprietary
13	8◄─►22	gemini-2.5-flash	1213	±7	40,526	Google	Proprietary
14	8◄─►24	gpt-5-high	1210	±8	32,822	OpenAI	Proprietary
15	8◄─►28	claude-opus-4-20250514-thinking-16k	1209	±15	1,416	Anthropic	Proprietary
16	8◄─►26	qwen3-vl-235b-a22b-instruct	1209	±10	5,408	Alibaba	Apache 2.0
17	8◄─►28	ernie-5.0-preview-1120	1206	±12	2,844	Baidu	Proprietary
18	8◄─►29	claude-sonnet-4-20250514-thinking-32k	1204	±16	1,307	Anthropic	Proprietary
19	12◄─►28	gpt-4.1-mini-2025-04-14	1202	±8	39,483	OpenAI	Proprietary
20	12◄─►28	o4-mini-2025-04-16	1202	±8	39,975	OpenAI	Proprietary
21	12◄─►33	claude-3-7-sonnet-20250219-thinking-32k	1195	±15	1,627	Anthropic	Proprietary
22	13◄─►33	hunyuan-vision-1.5-thinking	1193	±12	2,731	Tencent	Proprietary
23	14◄─►33	o1-2024-12-17	1192	±10	3,694	OpenAI	Proprietary
24	14◄─►33	qwen3-vl-235b-a22b-thinking	1190	±12	2,524	Alibaba	Apache 2.0
25	16◄─►33	gemini-2.5-flash-lite-preview-06-17-thinking	1188	±8	36,712	Google	Proprietary
26	15◄─►33	claude-opus-4-20250514	1188	±12	2,512	Anthropic	Proprietary
27	15◄─►34	claude-sonnet-4-20250514	1186	±13	2,032	Anthropic	Proprietary
28	20◄─►34	gpt-5-mini-high	1184	±9	28,883	OpenAI	Proprietary
29	16◄─►34	qwen-vl-max-2025-08-13	1184	±12	3,314	Alibaba	Proprietary
30	21◄─►34	grok-4-0709	1180	±8	31,494	xAI	Proprietary
31	21◄─►38	gemini-2.5-flash-lite-preview-09-2025-no-thinking	1178	±10	5,085	Google	Proprietary
32	21◄─►34	gemini-1.5-pro-002	1178	±8	8,916	Google	Proprietary
33	21◄─►38	claude-3-7-sonnet-20250219	1177	±9	4,659	Anthropic	Proprietary
34	27◄─►41	gemini-2.0-flash-001	1168	±7	9,735	Google	Proprietary
35	32◄─►45	gpt-4o-2024-05-13	1162	±8	23,280	OpenAI	Proprietary
36	32◄─►45	claude-3-5-sonnet-20241022	1161	±7	10,527	Anthropic	Proprietary
37	32◄─►49	glm-4.5v	1157	±12	3,478	Z.ai	MIT
38	34◄─►49	mistral-medium-2505	1156	±9	11,429	Mistral	Proprietary
39	34◄─►49	gemma-3-27b-it	1155	±8	18,266	Google	Gemma
40	32◄─►50	step-1o-turbo-202506	1154	±14	2,054	StepFun	Proprietary
41	35◄─►50	mistral-medium-2508	1149	±8	33,624	Mistral	Proprietary
42	34◄─►55	hunyuan-large-vision	1148	±16	1,459	Tencent	Proprietary
43	35◄─►50	claude-3-5-sonnet-20240620	1147	±9	21,632	Anthropic	Proprietary
44	35◄─►52	gpt-5-nano-high	1146	±12	4,102	OpenAI	Proprietary
45	35◄─►50	llama-4-maverick-17b-128e-instruct	1146	±9	7,158	Meta	Llama 4
46	37◄─►56	step-3	1141	±12	3,440	StepFun	Apache 2.0
47	37◄─►56	mistral-small-2506	1139	±10	11,645	Mistral	Apache 2.0
48	37◄─►56	gemini-1.5-flash-002	1138	±9	7,256	Google	Proprietary
49	37◄─►58	claude-3-5-haiku-20241022	1135	±15	1,564	Anthropic	Proprietary
50	40◄─►58	gemini-2.0-flash-lite-preview-02-05	1133	±10	4,001	Google	Proprietary
51	45◄─►58	mistral-small-3.1-24b-instruct-2503	1125	±9	27,478	Mistral	Apache 2.0
52	44◄─►58	llama-4-scout-17b-16e-instruct	1125	±10	6,627	Meta	Llama
53	44◄─►58	step-1o-vision-32k-highres	1123	±12	2,844	StepFun	Proprietary
54	45◄─►58	qwen2.5-vl-72b-instruct	1121	±10	3,767	Alibaba	Qwen
55	46◄─►59	gpt-4o-2024-08-06	1118	±12	3,384	OpenAI	Proprietary
56	49◄─►59	gemini-1.5-pro-001	1117	±11	16,735	Google	Proprietary
57	45◄─►61	qwen2.5-vl-32b-instruct	1117	±15	1,490	Alibaba	Apache 2.0
58	49◄─►61	gpt-4-turbo-2024-04-09	1112	±11	13,390	OpenAI	Proprietary
59	57◄─►63	gpt-4o-mini-2024-07-18	1097	±7	17,362	OpenAI	Proprietary
60	57◄─►63	pixtral-large-2411	1094	±9	5,439	Mistral	MRL
61	55◄─►67	gpt-4.1-nano-2025-04-14	1088	±18	1,211	OpenAI	Proprietary
62	59◄─►65	qwen2-vl-72b	1085	±9	5,946	Alibaba	Qwen
63	59◄─►67	qwen-vl-max-1119	1085	±16	1,431	Alibaba	Proprietary
64	61◄─►69	gemini-1.5-flash-8b-001	1070	±10	6,255	Google	Proprietary
65	62◄─►69	claude-3-opus-20240229	1064	±10	15,568	Anthropic	Proprietary
66	61◄─►69	step-1v-32k	1061	±16	1,551	StepFun	Proprietary
67	62◄─►69	gemini-1.5-flash-001	1059	±11	13,261	Google	Proprietary
68	64◄─►72	molmo-72b-0924	1047	±12	3,063	AI2	Apache 2.0
69	64◄─►76	hunyuan-standard-vision-2024-12-31	1043	±21	809	Tencent	Proprietary
70	68◄─►76	llama-3.2-vision-90b-instruct	1032	±8	8,708	Meta	Llama 3.2
71	68◄─►78	qwen2-vl-7b-instruct	1031	±10	5,778	Aliaba	Apache 2.0
72	69◄─►78	pixtral-12b-2409	1025	±9	7,520	Mistral	Apache 2.0
73	68◄─►79	internvl2-26b	1024	±12	5,146	OpenGVLab	MIT
74	69◄─►81	amazon-nova-lite-v1.0	1020	±15	1,856	Amazon	Proprietary
75	69◄─►81	amazon-nova-pro-v1.0	1020	±13	2,338	Amazon	Proprietary
76	69◄─►80	claude-3-sonnet-20240229	1019	±11	12,314	Anthropic	Proprietary
77	71◄─►83	yi-vision	1004	±18	1,220	01 AI	Proprietary
78	73◄─►83	claude-3-haiku-20240307	1002	±12	13,380	Anthropic	Proprietary
79	71◄─►85	c4ai-aya-vision-32b	1000	±22	847	Cohere	CC-BY-NC-4.0
80	74◄─►83	molmo-7b-d-0924	996	±13	2,828	AI2	Apache 2.0
81	77◄─►83	llama-3.2-vision-11b-instruct	991	±11	4,818	Meta	Llama 3.2
82	75◄─►87	nvila-internal-15b-v1	987	±20	1,078	NVIDIA	-
83	77◄─►87	llava-onevision-qwen2-72b-ov	980	±18	1,321	LLaVA	Apache 2.0
84	82◄─►87	llava-v1.6-34b	966	±12	4,531	LLaVA	Apache 2.0
85	81◄─►87	minicpm-v-2_6	964	±15	1,987	OpenBMB	Apache 2.0
86	81◄─►87	cogvlm2-llama3-chat-19b	964	±15	1,991	Zhipu AI	CogVLM2
87	82◄─►87	internvl2-4b	957	±12	3,706	OpenGVLab	MIT
88	88◄─►88	phi-3.5-vision-instruct	922	±15	2,592	Microsoft	MIT
89	89◄─►89	phi-3-vision-128k-instruct	883	±18	1,401	Microsoft	MIT

Vision Arena

Default Leaderboard Plots

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)

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