llm leaderboard

Best LLM for vision

Vision covers reading a document, following a diagram, and describing what is in a frame.

Aldena runs these models inside your team's projects. See what each one costs.

142 of 142 ranked models
Ranked models
rankmodelvendorcompositepricein / outbenchmarkseloeloeloeloeloelo
1
Claude Fable 5 (high)anthropic/claude-fable-5:high
Anthropic83.8$10.00 / $50.004 of 6 benchmarks1309132713501353
2
Claude Opus 5.5 (high)anthropic/claude-opus-5.5:high
Anthropic82.1$4.00 / $20.004 of 6 benchmarks1295131213381347
3
Gemini 3.8 Flash (high)google/gemini-3.8-flash:high
Google81.9$0.75 / $3.754 of 6 benchmarks1293131513411339
4
Claude Opus 4.7 (high)anthropic/claude-opus-4.7:high
Anthropic81.8$5.00 / $25.005 of 6 benchmarks12981311133413311246
5
Qwen3.8 Maxqwen/qwen3.8-max
Qwen81.2—4 of 6 benchmarks1299131613221334
6
Claude Opus 4.6 (high)anthropic/claude-opus-4.6:high
Anthropic79.6$5.00 / $25.005 of 6 benchmarks12991313132413241237
7
GPT-6.1 Sol (max)openai/gpt-6.1-sol:max
OpenAI79.2$2.00 / $10.004 of 6 benchmarks1292129913271354
8
Claude Opus 4.6anthropic/claude-opus-4.6
Anthropic78.4$5.00 / $25.005 of 6 benchmarks12931311133013111239
9
Gemini 3 Progoogle/gemini-3-pro
Google78.3—6 of 6 benchmarks128913021302130812901277
10
Claude Fable 5.1 (max)anthropic/claude-fable-5.1:max
Anthropic78.2$10.00 / $50.004 of 6 benchmarks1289130413281328
11
Gemini 3.1 Pro Previewgoogle/gemini-3.1-pro-preview
Google77.2$2.00 / $12.006 of 6 benchmarks128012951309131112841280
12
Gemini 3.7 Flash (high)google/gemini-3.7-flash:high
Google77.2$0.75 / $3.754 of 6 benchmarks1294130813201313
13
Claude Opus 4.8 (high)anthropic/claude-opus-4.8:high
Anthropic76.3$5.00 / $25.004 of 6 benchmarks1286129913171331
14
Claude Opus 5 (high)anthropic/claude-opus-5:high
Anthropic76.3$5.00 / $25.004 of 6 benchmarks1285130013261321
15
Muse Spark 1.3 (max)meta/muse-spark-1.3:max
Meta75.1$1.25 / $4.254 of 6 benchmarks1291130813291301
16
GPT-6 Astra (max)openai/gpt-6-astra:max
OpenAI75.0$10.00 / $50.004 of 6 benchmarks1284129813111341
17
Muse Spark 1.2 (xhigh)meta/muse-spark-1.2:xhigh
Meta74.9$1.25 / $4.253 of 6 benchmarks129413061316
18
GPT-5.6 Sol (xhigh)openai/gpt-5.6-sol:xhigh
OpenAI74.8$4.00 / $20.004 of 6 benchmarks1288129813181316
19
Muse Spark 1.3 (xhigh)meta/muse-spark-1.3:xhigh
Meta74.3$1.25 / $4.254 of 6 benchmarks1288130113151311
20
GPT-5.5 (high)openai/gpt-5.5:high
OpenAI74.1$5.00 / $30.005 of 6 benchmarks12831298131613351223
21
Claude Opus 4.7anthropic/claude-opus-4.7
Anthropic73.4$5.00 / $25.005 of 6 benchmarks13001315133013251183
22
Muse Sparkmeta/muse-spark
Meta73.0—4 of 6 benchmarks1294130313121299
23
Gemini 3 Flash Previewgoogle/gemini-3-flash-preview
Google72.4$0.50 / $3.006 of 6 benchmarks127212851293130912921238
24
Gemini 3.5 Flash (medium)google/gemini-3.5-flash:medium
Google72.2$1.50 / $9.004 of 6 benchmarks1281129813101314
25
GPT-5.5openai/gpt-5.5
OpenAI71.9$5.00 / $30.005 of 6 benchmarks12851296132113301213
26
Gemini 3.6 Flash (high)google/gemini-3.6-flash:high
Google71.2$0.75 / $3.754 of 6 benchmarks1280129813121308
27
Claude Opus 4.8anthropic/claude-opus-4.8
Anthropic70.3$5.00 / $25.004 of 6 benchmarks1279129313061321
28
GLM 5.3 Flashz-ai/glm-5.3-flash
Z.ai69.6$0.15 / $0.505 of 6 benchmarks12721287131013121237
29
GPT-5.4 (high)openai/gpt-5.4:high
OpenAI69.4$2.50 / $15.005 of 6 benchmarks12821297131913301193
30
Gemini 3.5 Flash (high)google/gemini-3.5-flash:high
Google68.9$1.50 / $9.004 of 6 benchmarks1281129113071307
31
GPT-5.2 Chatopenai/gpt-5.2-chat
OpenAI68.7—5 of 6 benchmarks12781288130712931241
32
Grok 4.5x-ai/grok-4.5
SpaceXAI68.1$2.00 / $6.004 of 6 benchmarks1280129013071305
33
Claude Sonnet 5.5 (xhigh)anthropic/claude-sonnet-5.5:xhigh
Anthropic67.5$2.00 / $10.004 of 6 benchmarks1263128113061319
34
GPT-5.6 Terra (xhigh)openai/gpt-5.6-terra:xhigh
OpenAI66.3$2.00 / $12.004 of 6 benchmarks1266127712961318
35
Gemini 3 Flash Preview (minimal)google/gemini-3-flash-preview:minimal
Google65.8$0.50 / $3.006 of 6 benchmarks126012721285129712741212
36
GPT-5.4openai/gpt-5.4
OpenAI65.7$2.50 / $15.005 of 6 benchmarks12781290131313041213
37
Muse Spark 1.1meta/muse-spark-1.1
Meta65.3$1.25 / $4.254 of 6 benchmarks1279129413051285
38
GPT 5.5 Instantopenai/gpt-5.5-instant
OpenAI64.4—4 of 6 benchmarks1277128713101279
39
GPT-6 Sol (max)openai/gpt-6-sol:max
OpenAI64.3$2.00 / $10.004 of 6 benchmarks1258127812971310
40
Kimi K2.6moonshotai/kimi-k2.6
MoonshotAI63.2$0.65 / $3.414 of 6 benchmarks1265128012931303
41
DeepSeek V4.1 Flash (max)deepseek/deepseek-v4.1-flash:max
DeepSeek63.1$0.051 / $0.5674 of 6 benchmarks1263127412961307
42
Claude Sonnet 5 (high)anthropic/claude-sonnet-5:high
Anthropic62.4$2.00 / $10.004 of 6 benchmarks1261127912921303
43
Gemini 2.5 Progoogle/gemini-2.5-pro
Google61.6$1.25 / $10.006 of 6 benchmarks124712591267127312511251
44
Gemini 3.5 Flash Litegoogle/gemini-3.5-flash-lite
Google61.5$0.30 / $2.504 of 6 benchmarks1266128012871288
45
GPT-5.1 (high)openai/gpt-5.1:high
OpenAI61.4$1.25 / $10.006 of 6 benchmarks125112621277128512291249
46
Qwen3.7 Plusqwen/qwen3.7-plus
Qwen61.2$0.32 / $1.284 of 6 benchmarks1262128012881288
47
Grok 4.6 (high)x-ai/grok-4.6:high
SpaceXAI61.0$2.00 / $6.004 of 6 benchmarks1263127512951288
48
Kimi K2.5 (thinking)moonshotai/kimi-k2.5:thinking
MoonshotAI59.2$0.532 / $3.3256 of 6 benchmarks125212661280128912361193
49
Grok 4.7 (xhigh)x-ai/grok-4.7:xhigh
SpaceXAI59.2$2.00 / $6.004 of 6 benchmarks1249126212841307
50
Grok 4.20 Beta 0309 (reasoning)x-ai/grok-4.20-beta-0309:reasoning
SpaceXAI58.9—5 of 6 benchmarks12561267127512661240
51
Claude Sonnet 4.6anthropic/claude-sonnet-4.6
Anthropic58.8$3.00 / $15.005 of 6 benchmarks12751290130513021168
52
Dola Seed 2.0 Probytedance/dola-seed-2.0-pro
ByteDance57.9—4 of 6 benchmarks1256127012881277
53
GPT-5.6 Luna (xhigh)openai/gpt-5.6-luna:xhigh
OpenAI57.6$0.20 / $1.204 of 6 benchmarks1255126512781289
54
GPT-5.2 (high)openai/gpt-5.2:high
OpenAI57.3$1.75 / $14.006 of 6 benchmarks124312601276128112021261
55
Gemini 3.1 Flash Lite Previewgoogle/gemini-3.1-flash-lite-preview
Google57.2$0.25 / $1.506 of 6 benchmarks123512471252127112691218
56
GPT-5.4 Mini (high)openai/gpt-5.4-mini:high
OpenAI55.7$0.75 / $4.505 of 6 benchmarks12501263128312911204
57
Grok 4.20 Multi Agent Beta 0309x-ai/grok-4.20-multi-agent-beta-0309
SpaceXAI55.3—5 of 6 benchmarks12511260126712741228
58
Step 5 Previewstepfun/step-5-preview
StepFun55.2$1.00 / $2.704 of 6 benchmarks1246126612881269
59
Gemma 4 31Bgoogle/gemma-4-31b-it
Google54.5$0.20 / $0.406 of 6 benchmarks126512771287130411981161
60
MiMo-V2.6-Flashxiaomi/mimo-v2.6-flash
Xiaomi54.0$0.14 / $0.284 of 6 benchmarks1248126512651283
61
Qwen3.8 27Bqwen/qwen3.8-27b
Qwen53.1$0.15 / $1.8754 of 6 benchmarks1244126212671275
62
Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b
Qwen52.1$0.60 / $3.606 of 6 benchmarks124612601273128712201171
63
GPT-5 (high)openai/gpt-5:high
OpenAI51.3$1.25 / $10.006 of 6 benchmarks121112321251125912611195
64
MiMo-V2.6-Proxiaomi/mimo-v2.6-pro
Xiaomi51.0$0.435 / $0.874 of 6 benchmarks1247125912721256
65
Chatgpt 4oopenai/chatgpt-4o
OpenAI50.6—6 of 6 benchmarks123812461260124512181215
66
Gemma 4 26B A4B google/gemma-4-26b-a4b-it
Google50.1$0.0765 / $0.2556 of 6 benchmarks124212571261128511591229
67
GPT 4.5 Previewopenai/gpt-4.5-preview
OpenAI49.9—1 of 6 benchmarks1225
68
Grok 4.3x-ai/grok-4.3
SpaceXAI49.7$1.25 / $2.505 of 6 benchmarks12421251126712551214
69
GPT-6 Luna (max)openai/gpt-6-luna:max
OpenAI49.6$0.10 / $0.504 of 6 benchmarks1236124812671271
70
GPT 5 Chatopenai/gpt-5-chat
OpenAI47.7—6 of 6 benchmarks122512451258127511891195
71
Mistral Large 4mistralai/mistral-large-4-0
Mistral46.5$0.70 / $2.104 of 6 benchmarks1230124212651256
72
o1openai/o1
OpenAI45.4—1 of 6 benchmarks1193
73
Gemini 2.5 Flashgoogle/gemini-2.5-flash
Google45.0$0.30 / $2.506 of 6 benchmarks121512231233123912211210
74
o3openai/o3
OpenAI44.9$2.00 / $8.006 of 6 benchmarks121512261233124912311184
75
Kimi K2.5 Instantmoonshotai/kimi-k2.5-instant
Moonshot AI44.7—4 of 6 benchmarks1236124412511248
76
Qwen3 VL 235B A22B Instructqwen/qwen3-vl-235b-a22b-instruct
Qwen43.7$0.21 / $1.906 of 6 benchmarks121512271238125811761212
77
GPT-5.1openai/gpt-5.1
OpenAI43.7$1.25 / $10.006 of 6 benchmarks124012521262125711881165
78
Qwen3.5-122B-A10Bqwen/qwen3.5-122b-a10b
Qwen43.6$0.26 / $2.084 of 6 benchmarks1226123912441254
79
MiMo-V2.5xiaomi/mimo-v2.5
Xiaomi43.5$0.40 / $2.005 of 6 benchmarks12341248126712541169
80
Gemini 1.5 Pro 002google/gemini-1.5-pro-002
Google43.2—1 of 6 benchmarks1180
81
Mimo v2 Omnixiaomi/mimo-v2-omni
Xiaomi42.7—4 of 6 benchmarks1216123112551248
82
GPT-4.1openai/gpt-4.1
OpenAI42.3$2.00 / $8.006 of 6 benchmarks121412271232125211841211
83
GPT-4o (2024-05-13)openai/gpt-4o-2024-05-13
OpenAI42.3$5.00 / $15.001 of 6 benchmarks1162
84
GLM 5V Turboz-ai/glm-5v-turbo
Z.ai42.1$1.20 / $4.006 of 6 benchmarks123212451264125211741173
85
o4 Miniopenai/o4-mini
OpenAI41.9—6 of 6 benchmarks120112111219125412211184
86
MiniMax M3minimax/minimax-m3
MiniMax41.8$0.30 / $1.205 of 6 benchmarks12321244126012581160
87
GPT-5.2openai/gpt-5.2
OpenAI40.5$1.75 / $14.006 of 6 benchmarks122812381257126411721166
88
Ernie 5.0 Preview 1220baidu/ernie-5.0-preview-1220
Baidu39.9—4 of 6 benchmarks1219123012221243
89
Claude Sonnet 4 (thinking 32K)anthropic/claude-sonnet-4:thinking-32k
Anthropic39.3—3 of 6 benchmarks120612161228
90
Gemini 1.5 Flash 002google/gemini-1.5-flash-002
Google39.2—1 of 6 benchmarks1142
91
Inkling Smallthinkingmachines/inkling-small
Thinking Machines38.8$0.45 / $1.204 of 6 benchmarks1208122212401239
92
Gemini 2.0 Flash Lite Preview 02 05google/gemini-2.0-flash-lite-preview-02-05
Google38.8—1 of 6 benchmarks1136
93
Qwen VL (max)qwen/qwen-vl:max
Qwen38.6—3 of 6 benchmarks118512131248
94
Nemotron 3.5 Super VLnvidia/nemotron-3.5-super-vl
NVIDIA38.6$0.15 / $0.474 of 6 benchmarks1212122412391233
95
Qwen2.5 VL 72B Instructqwen/qwen2.5-vl-72b-instruct
Qwen37.8—1 of 6 benchmarks1121
96
Gemini 1.5 Pro 001google/gemini-1.5-pro-001
Google37.6—1 of 6 benchmarks1120
97
Qwen2.5 VL 32B Instructqwen/qwen2.5-vl-32b-instruct
Qwen37.3—1 of 6 benchmarks1118
98
GPT-4o (2024-08-06)openai/gpt-4o-2024-08-06
OpenAI37.1$2.50 / $10.001 of 6 benchmarks1118
99
GPT-4 Turboopenai/gpt-4-turbo
OpenAI36.6—1 of 6 benchmarks1112
100
GPT-4.1 Miniopenai/gpt-4.1-mini
OpenAI36.5$0.40 / $1.606 of 6 benchmarks120112071208122111971193
101
Mistral Medium 3.5mistralai/mistral-medium-3-5
Mistral36.5$1.50 / $7.504 of 6 benchmarks1200121512431216
102
GPT-4o-mini (2024-07-18)openai/gpt-4o-mini-2024-07-18
OpenAI36.4$0.15 / $0.601 of 6 benchmarks1097
103
GPT-4.1 Nanoopenai/gpt-4.1-nano
OpenAI36.1$0.10 / $0.401 of 6 benchmarks1089
104
GPT-5 Mini (high)openai/gpt-5-mini:high
OpenAI36.1$0.25 / $2.006 of 6 benchmarks118512001213123812121173
105
Gemini 1.5 Flash 8B 001google/gemini-1.5-flash-8b-001
Google35.9—1 of 6 benchmarks1071
106
Claude 3 Opusanthropic/claude-3-opus
Anthropic35.7—1 of 6 benchmarks1062
107
Grok 4.1 Fast (reasoning)x-ai/grok-4-1-fast:reasoning
SpaceXAI35.4—5 of 6 benchmarks11961194121811711220
108
Gemini 1.5 Flash 001google/gemini-1.5-flash-001
Google35.4—1 of 6 benchmarks1059
109
Amazon Nova Lite v1.0amazon/amazon-nova-lite-v1.0
Amazon35.2—1 of 6 benchmarks1019
110
Mistral Large 3mistralai/mistral-large-3
Mistral AI35.1—4 of 6 benchmarks1200121412331203
111
Amazon Nova Pro v1.0amazon/amazon-nova-pro-v1.0
Amazon35.0—1 of 6 benchmarks1019
112
Claude Opus 4 (thinking 16K)anthropic/claude-opus-4:thinking-16k
Anthropic34.9—4 of 6 benchmarks1208121912091213
113
Claude 3 Sonnetanthropic/claude-3-sonnet
Anthropic34.7—1 of 6 benchmarks1017
114
Claude 3 Haikuanthropic/claude-3-haiku
Anthropic34.5—1 of 6 benchmarks1000
115
Grok 4 0709x-ai/grok-4-0709
SpaceXAI34.3—6 of 6 benchmarks118411761192117112361171
116
Claude Opus 4anthropic/claude-opus-4
Anthropic34.0—4 of 6 benchmarks1189119812061243
117
Claude 3.7 Sonnet (thinking 32K)anthropic/claude-3-7-sonnet:thinking-32k
Anthropic33.2—4 of 6 benchmarks1193121012131214
118
Qwen3 VL 235B A22B Thinkingqwen/qwen3-vl-235b-a22b-thinking
Qwen32.6—4 of 6 benchmarks1188119912081226
119
Qwen3.5-27Bqwen/qwen3.5-27b
Qwen32.4$0.26 / $2.606 of 6 benchmarks121612271241124211511091
120
Claude Sonnet 4anthropic/claude-sonnet-4
Anthropic31.7—4 of 6 benchmarks1189119112021226
121
Gemini 2.5 Flash Lite (thinking)google/gemini-2.5-flash-lite:thinking
Google31.6$0.10 / $0.406 of 6 benchmarks118911901188119411831183
122
Claude 3.7 Sonnetanthropic/claude-3-7-sonnet
Anthropic31.0—4 of 6 benchmarks1175118711941230
123
GPT-5.4 Nano (high)openai/gpt-5.4-nano:high
OpenAI30.5$0.20 / $1.255 of 6 benchmarks11991213122612331117
124
Gemini 2.5 Flash Lite (nothinking)google/gemini-2.5-flash-lite:nothinking
Google28.6$0.10 / $0.404 of 6 benchmarks1174118511821198
125
Hunyuan Vision 1.5 (thinking)tencent/hunyuan-vision-1.5:thinking
Tencent27.5—4 of 6 benchmarks1154115811841227
126
Claude 3.5 Sonnetanthropic/claude-3-5-sonnet
Anthropic27.1—4 of 6 benchmarks1161118111861171
127
GLM 4.6Vz-ai/glm-4.6v
Z.ai26.0$0.30 / $0.904 of 6 benchmarks1161117211901155
128
Gemini 2.0 Flash 001google/gemini-2.0-flash-001
Google26.0—5 of 6 benchmarks11721167117711941150
129
Hunyuan Large Visiontencent/hunyuan-large-vision
Tencent24.9—3 of 6 benchmarks115011461116
130
GPT-5 Nano (high)openai/gpt-5-nano:high
OpenAI24.7$0.05 / $0.404 of 6 benchmarks1147115911541188
131
GLM 4.5Vz-ai/glm-4.5v
Z.ai23.8$0.60 / $1.804 of 6 benchmarks1153115811661165
132
Step 3stepfun/step-3
StepFun23.1—4 of 6 benchmarks1144115011451174
133
Llama 4 Maverick 17B 128e Instructmeta-llama/llama-4-maverick-17b-128e-instruct
Meta22.8—4 of 6 benchmarks1146115411521166
134
Step 1o Turbo 202506stepfun/step-1o-turbo-202506
StepFun22.8—4 of 6 benchmarks1155115111621130
135
Gemma 3 27Bgoogle/gemma-3-27b-it
Google22.4$0.08 / $0.456 of 6 benchmarks116011621165116911701119
136
Molmo 2 8Ballenai/molmo-2-8b
Allen Institute for AI22.3—3 of 6 benchmarks111310941118
137
Mistral Medium 2508mistralai/mistral-medium-2508
Mistral AI21.9—6 of 6 benchmarks116011731180117411291128
138
Llama 4 Scout 17B 16e Instructmeta-llama/llama-4-scout-17b-16e-instruct
Meta20.5—4 of 6 benchmarks1128113111521151
139
Claude 3.5 Haikuanthropic/claude-3-5-haiku
Anthropic20.0—4 of 6 benchmarks1127113211311148
140
Mistral Medium 2505mistralai/mistral-medium-2505
Mistral AI19.2—6 of 6 benchmarks115511601167116711251091
141
Mistral Small 2506mistralai/mistral-small-2506
Mistral AI17.9—6 of 6 benchmarks114011411140117811341053
142
Mistral Small 3.1 24B Instruct 2503mistralai/mistral-small-3.1-24b-instruct-2503
Mistral AI17.8—6 of 6 benchmarks112711341129116211301133
How this ranks

Every benchmark value becomes a percentile among the models that have it, so accuracy scores, Elo ratings and word error rates compare without hand-tuned scaling. Metrics where lower is better are inverted first. Raw values are never summed or averaged across benchmarks. A model's mean percentile is then shrunk toward the mean of the models that were broadly benchmarked, so a model tested twice cannot outrank a broadly tested one on two lucky results. Turning a data source off runs that same ranking code again in your browser over the sources you left on.

A model scored on fewer than 3 of the 6 ranked benchmarks in this category still ranks here, on the benchmarks it does have, and its row carries a partial coverage mark. On an equal score it sits under the model that earned the same number across more of the board.

Data sources

Turn a source off to drop every benchmark it feeds and re-rank the board in your browser. Your choice follows you across the leaderboard pages.

  • LMArenaCC BY 4.0

    Arena ratings by LMArena, from the public leaderboard dataset.