Vision Models
Updated
Models that understand images as well as text: photos, diagrams, screenshots and handwriting.
- Best overall
Highest ToolsGuide Score with images in the prompt
639 / 1,000
Vision model leaderboard
167 models · ToolsGuide Score out of 1,000, where 1,000 is a flawless model
20 of 167 models in the chart · ·
All models
| Compare | # | Model | Creator | Licence | Input / 1M | Score | From votes | From tests · count | Votes |
|---|---|---|---|---|---|---|---|---|---|
| 1 | GPT-6 Astra (Max) |   OpenAI | Proprietary | $10.00 | 639 | 548 | 721 · 3 | 4,012 | |
| 2 | GPT-5.5 | OpenAI | Proprietary | $5.00 | 600 | 552 | 648 · 4 | 28,781 | |
| 3 | GPT-5.5 (High) | OpenAI | Proprietary | $5.00 | 594 | 539 | 648 · 4 | 27,067 | |
| 4 | Gemini 3.1 Pro Preview |  Google | Proprietary | $2.00 | 587 | 530 | 643 · 3 | 47,114 | |
| 5 | Claude Opus 5.5 (High) |   Anthropic | Proprietary | $4.00 | 574 | 574 | – | 3,035 | |
| 6 | Muse Spark 1.2 (xHigh) |  Meta | Proprietary | $0.10 | 573 | 573 | – | 1,948 | |
| 7 | Muse Spark 1.3 (Max) | Meta | Proprietary | $0.10 | 563 | 563 | – | 5,392 | |
| 8 | Claude Opus 4.7 | Anthropic | Proprietary | $5.00 | 560 | 593 | 493 · 2 | 23,139 | |
| 9 | Claude Opus 5 (High) | Anthropic | Proprietary | $5.00 | 558 | 558 | – | 17,743 | |
| 10 | Claude Opus 4.7 (High) | Anthropic | Proprietary | $5.00 | 557 | 590 | 493 · 2 | 22,731 | |
| 11 | Muse Spark 1.3 (xHigh) | Meta | Proprietary | $0.10 | 557 | 557 | – | 4,897 | |
| 12 | Claude Fable 5.1 (Max) | Anthropic | Proprietary | $10.00 | 557 | 557 | – | 5,026 | |
| 13 | Claude Opus 4.6 (High) | Anthropic | Proprietary | $5.00 | 555 | 594 | 517 · 4 | 22,286 | |
| 14 | Claude Fable 5 (High) | Anthropic | Proprietary | $10.00 | 554 | 626 | 420 · 2 | 14,714 | |
| 15 | Claude Opus 4.6 | Anthropic | Proprietary | $5.00 | 547 | 577 | 517 · 4 | 26,841 | |
| 16 | Gemini 3.5 Flash (Medium) | Google | Proprietary | $1.50 | 539 | 540 | 539 · 4 | 14,061 | |
| 17 | Gemini 3.5 Flash (High) | Google | Proprietary | $1.50 | 539 | 539 | 539 · 4 | 13,862 | |
| 18 | Gemini 3.6 Flash (High) | Google | Proprietary | $1.50 | 538 | 538 | – | 8,979 | |
| 19 | Claude Sonnet 4.6 | Anthropic | Proprietary | $3.00 | 535 | 516 | 554 · 3 | 27,408 | |
| 20 | Grok 4.5 |   xAI | Proprietary | $2.00 | 529 | 529 | – | 12,599 | |
| 21 | GPT-5.2 Chat (20260210) | OpenAI | Proprietary | – | 524 | 524 | – | 16,465 | |
| 22 | Claude Sonnet 5 (High) | Anthropic | Proprietary | $2.00 | 518 | 475 | 562 · 3 | 14,034 | |
| 23 | Claude Opus 4.8 (High) | Anthropic | Proprietary | $5.00 | 515 | 551 | 481 · 3 | 18,303 | |
| 24 | GPT-6.1 Sol (Max) | OpenAI | Proprietary | $2.00 | 510 | 576 | 394 · 2 | 2,330 | |
| 25 | Claude Opus 4.8 | Anthropic | Proprietary | $5.00 | 504 | 528 | 481 · 3 | 18,800 | |
| 26 | Gemini 3.7 Flash (High) | Google | Proprietary | $0.75 | 494 | 582 | 416 · 3 | 5,369 | |
| 27 | GPT-5.6 Sol (xHigh) | OpenAI | Proprietary | $5.00 | 491 | 552 | 387 · 2 | 11,439 | |
| 28 | Gemini 3.5 Flash-Lite | Google | Proprietary | $0.30 | 490 | 490 | – | 9,238 | |
| 29 | Qwen3.8-Max |  Alibaba | Proprietary | $1.65 | 490 | 598 | 397 · 7 | 9,957 | |
| 30 | Gemini 3.8 Flash (High) | Google | Proprietary | $0.75 | 488 | 568 | 416 · 3 | 4,645 | |
| 31 | GPT-6 Sol (Max) | OpenAI | Proprietary | $2.00 | 488 | 488 | – | 3,221 | |
| 32 | Claude Sonnet 5.5 (xHigh) | Anthropic | Proprietary | $2.00 | 484 | 484 | – | 2,564 | |
| 33 | Muse Spark | Meta | Proprietary | – | 483 | 578 | 401 · 9 | 5,865 | |
| 34 | Grok 4.6 (High) | xAI | Proprietary | $2.00 | 476 | 476 | – | 7,261 | |
| 35 | DeepSeek-V4.1-Flash-20260910 (Max) |  DeepSeek | MIT | – | 473 | 473 | – | 4,100 | |
| 36 | Muse Spark 1.1 | Meta | Proprietary | $1.25 | 465 | 532 | 406 · 3 | 12,804 | |
| 37 | GPT-5.4 (High) | OpenAI | Proprietary | $2.50 | 460 | 545 | 387 · 5 | 30,747 | |
| 38 | Dola Seed 2.0 Pro |  ByteDance | Proprietary | – | 459 | 459 | – | 10,972 | |
| 39 | Grok 4.20 Beta (Reasoning, 0309) | xAI | Proprietary | – | 455 | 455 | – | 26,796 | |
| 40 | GPT-5.4 | OpenAI | Proprietary | $2.50 | 452 | 528 | 387 · 5 | 22,804 | |
| 41 | GPT-5.6 Terra (xHigh) | OpenAI | Proprietary | $2.00 | 446 | 494 | 365 · 2 | 11,842 | |
| 42 | Mistral Large 4 |  Mistral | Proprietary | $0.70 | 446 | 446 | – | 1,008 | |
| 43 | GPT-5.1 (High) | OpenAI | Proprietary | $1.25 | 444 | 444 | – | 9,848 | |
| 44 | Grok 4.20 Multi-Agent Beta (0309) | xAI | Proprietary | – | 441 | 441 | – | 24,532 | |
| 45 | GPT-5.5 Instant | OpenAI | Proprietary | $5.00 | 441 | 519 | 375 · 3 | 7,590 | |
| 46 | Grok 4.7 (xHigh) | xAI | Proprietary | $2.00 | 440 | 440 | – | 3,736 | |
| 47 | GPT-5.2 (High) | OpenAI | Proprietary | $1.75 | 437 | 420 | 455 · 7 | 19,022 | |
| 48 | GLM-5.3-Flash |   Z.ai | MIT | $0.15 | 433 | 504 | 374 · 4 | 16,254 | |
| 49 | Kimi K2.6 |   Moonshot AI | Modified MIT | $0.75 | 432 | 483 | 388 · 5 | 16,368 | |
| 50 | Step 5 Preview | StepFun | Proprietary | – | 432 | 432 | – | 1,649 | |
| 51 | Gemini 3 Pro | Google | Proprietary | – | 427 | 562 | 329 · 7 | 13,623 | |
| 52 | Qwen3.7-Plus | Alibaba | Proprietary | – | 427 | 479 | 381 · 15 | 14,212 | |
| 53 | MiMo-V2.6-Pro |  Xiaomi | MIT | $0.43 | 425 | 425 | – | 1,754 | |
| 54 | MiMo-V2.6-Flash | Xiaomi | MIT | $0.14 | 425 | 425 | – | 3,618 | |
| 55 | GPT-5.6 Luna (xHigh) | OpenAI | Proprietary | $0.20 | 421 | 459 | 357 · 2 | 11,055 | |
| 56 | Grok 4.3 | xAI | Proprietary | $1.25 | 419 | 419 | – | 27,901 | |
| 57 | GPT-5.2 | OpenAI | Proprietary | $1.75 | 417 | 384 | 455 · 7 | 20,100 | |
| 58 | GPT-5 (High) | OpenAI | Proprietary | – | 417 | 346 | 506 · 7 | 34,445 | |
| 59 | GPT-6 Luna (Max) | OpenAI | Proprietary | $0.10 | 414 | 414 | – | 3,662 | |
| 60 | GPT-5.1 | OpenAI | Proprietary | $1.25 | 413 | 413 | – | 10,801 | |
| 61 | ChatGPT-4o Latest (20250326) | OpenAI | Proprietary | – | 409 | 409 | – | 23,400 | |
| 62 | Kimi K2.5 (Thinking) | Moonshot AI | Modified MIT | – | 409 | 447 | 375 · 13 | 28,249 | |
| 63 | Gemma 4 31B | Google | Apache 2.0 | $0.09 | 407 | 475 | 352 · 4 | 40,219 | |
| 64 | GPT-5.4 mini (High) | OpenAI | Proprietary | $0.75 | 406 | 441 | 375 · 3 | 25,734 | |
| 65 | Gemini 3 Flash | Google | Proprietary | $0.50 | 406 | 508 | 329 · 7 | 39,131 | |
| 66 | Qwen3.5-397B-A17B | Alibaba | Apache 2.0 | $0.45 | 404 | 429 | 360 · 2 | 32,657 | |
| 67 | Qwen3.8-27B | Alibaba | Apache 2.0 | $0.40 | 402 | 419 | 386 · 7 | 7,614 | |
| 68 | Gemini 3 Flash (Minimal) | Google | Proprietary | $0.50 | 390 | 469 | 329 · 7 | 37,173 | |
| 69 | Kimi K2.5 (Instant) | Moonshot AI | Modified MIT | – | 389 | 404 | 375 · 13 | 4,111 | |
| 70 | MiniMax-M3 |  MiniMax | MiniMax Community License | $0.28 | 388 | 396 | 381 · 4 | 19,844 | |
| 71 | GLM-5V-Turbo | Z.ai | Proprietary | – | 388 | 388 | – | 42,757 | |
| 72 | Gemini 2.5 Flash Preview (09-2025) | Google | Proprietary | – | 388 | 388 | – | 4,792 | |
| 73 | MiMo-V2.5 | Xiaomi | MIT | $0.17 | 384 | 397 | 371 · 3 | 29,381 | |
| 74 | Inkling-Small | Thinky | Apache 2.0 | $0.30 | 379 | 342 | 423 · 5 | 8,074 | |
| 75 | GPT-5 Chat | OpenAI | Proprietary | – | 378 | 378 | – | 39,894 | |
| 76 | GPT-4.5 Preview (2025-02-27) | OpenAI | Proprietary | – | 376 | 376 | – | 2,925 | |
| 77 | Gemma 4 26B A4B | Google | Apache 2.0 | $0.07 | 376 | 418 | 341 · 4 | 23,922 | |
| 78 | Qwen3.5-122B-A10B | Alibaba | Apache 2.0 | $0.29 | 373 | 379 | 368 · 26 | 13,236 | |
| 79 | Gemini 3.1 Flash-Lite Preview | Google | Proprietary | $0.25 | 365 | 399 | 335 · 4 | 27,657 | |
| 80 | ERNIE-5.0-Preview (1220) |  Baidu | Proprietary | – | 364 | 364 | – | 3,349 | |
| 81 | Qwen3.5-27B | Alibaba | Apache 2.0 | $0.26 | 363 | 359 | 366 · 26 | 30,272 | |
| 82 | Gemini 2.5 Pro | Google | Proprietary | $1.25 | 361 | 434 | 306 · 4 | 88,755 | |
| 83 | MiMo-V2-Omni | Xiaomi | Proprietary | – | 357 | 357 | – | 10,650 | |
| 84 | GPT-5.4 nano (High) | OpenAI | Proprietary | $0.20 | 352 | 324 | 385 · 5 | 25,624 | |
| 85 | o3 (2025-04-16) | OpenAI | Proprietary | – | 351 | 357 | 345 · 9 | 45,567 | |
| 86 | Claude Opus 4 (Thinking 16K, 20250514) | Anthropic | Proprietary | – | 340 | 340 | – | 1,290 | |
| 87 | Claude Sonnet 4 (Thinking 32K, 20250514) | Anthropic | Proprietary | – | 338 | 338 | – | 1,144 | |
| 88 | Gemini 2.5 Flash | Google | Proprietary | $0.30 | 336 | 356 | 303 · 2 | 60,087 | |
| 89 | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B (NVFP4) |  NVIDIA | OpenMDW-1.1 | – | 333 | 333 | – | 2,509 | |
| 90 | Qwen3-VL-235B-A22B-Instruct | Alibaba | Apache 2.0 | $0.20 | 330 | 355 | 308 · 23 | 12,724 | |
| 91 | Mistral Medium 3.5 | Mistral | Modified MIT | $1.50 | 325 | 325 | – | 5,375 | |
| 92 | o4-mini (2025-04-16) | OpenAI | Proprietary | – | 325 | 328 | 323 · 4 | 41,316 | |
| 93 | Mistral Large 3 | Mistral | Apache 2.0 | $0.50 | 324 | 324 | – | 9,254 | |
| 94 | Grok 4 (0709) | xAI | Proprietary | – | 322 | 300 | 377 · 2 | 30,173 | |
| 95 | Grok 4.1 Fast (Reasoning) | xAI | Proprietary | $0.20 | 320 | 320 | – | 13,368 | |
| 96 | GPT-4.1 (2025-04-14) | OpenAI | Proprietary | $2.00 | 316 | 353 | 285 · 5 | 40,998 | |
| 97 | Claude 3.7 Sonnet (Thinking 32K, 20250219) | Anthropic | Proprietary | – | 314 | 314 | – | 1,436 | |
| 98 | Qwen3-VL-235B-A22B-Thinking | Alibaba | Apache 2.0 | – | 308 | 307 | 310 · 26 | 2,387 | |
| 99 | Claude Opus 4 (20250514) | Anthropic | Proprietary | – | 308 | 308 | – | 2,172 | |
| 99 | o1 (2024-12-17) | OpenAI | Proprietary | – | 308 | 314 | 296 · 2 | 3,694 | |
| 100 | Gemini 2.5 Flash-Lite Preview (Thinking, 06-17) | Google | Proprietary | – | 307 | 307 | – | 35,914 | |
| 101 | Claude Sonnet 4 (20250514) | Anthropic | Proprietary | – | 307 | 307 | – | 1,806 | |
| 102 | GPT-4.1 mini (2025-04-14) | OpenAI | Proprietary | $0.40 | 304 | 327 | 283 · 5 | 40,201 | |
| 103 | Qwen-VL-Max (2025-08-13) | Alibaba | Proprietary | – | 302 | 302 | – | 3,170 | |
| 104 | GPT-5 mini (High) | OpenAI | Proprietary | $0.25 | 300 | 300 | – | 28,943 | |
| 105 | Gemini 1.5 Pro (002) | Google | Proprietary | – | 288 | 293 | 283 · 3 | 8,902 | |
| 106 | Claude 3.7 Sonnet (20250219) | Anthropic | Proprietary | – | 287 | 287 | – | 4,517 | |
| 107 | Gemini 2.5 Flash-Lite Preview (Non-Thinking, 09-2025) | Google | Proprietary | – | 286 | 286 | – | 4,862 | |
| 108 | Gemini 2.0 Flash (001) | Google | Proprietary | – | 283 | 283 | – | 10,135 | |
| 109 | GPT-4o (2024-05-13) | OpenAI | Proprietary | $2.50 | 269 | 269 | – | 23,273 | |
| 110 | GLM-4.6V | Z.ai | MIT | – | 269 | 269 | – | 2,455 | |
| 111 | Claude 3.5 Sonnet (20241022) | Anthropic | Proprietary | – | 268 | 268 | – | 10,374 | |
| 112 | Mistral Medium 3.1 (2508) | Mistral | Proprietary | – | 267 | 267 | – | 41,611 | |
| 113 | Step-1o Turbo (202506) | StepFun | Proprietary | – | 262 | 262 | – | 1,895 | |
| 114 | Mistral Medium 3 (2505) | Mistral | Proprietary | – | 262 | 262 | – | 10,793 | |
| 115 | Hunyuan Vision 1.5 Thinking |  Tencent | Proprietary | – | 261 | 261 | – | 2,424 | |
| 116 | Gemini 1.5 Flash (002) | Google | Proprietary | – | 261 | 247 | 276 · 3 | 7,241 | |
| 117 | GLM-4.5V | Z.ai | MIT | – | 259 | 259 | – | 3,243 | |
| 118 | Hunyuan Large Vision | Tencent | Proprietary | – | 255 | 255 | – | 1,274 | |
| 119 | Gemma 3 27B IT | Google | Gemma | $0.08 | 254 | 267 | 243 · 6 | 16,982 | |
| 120 | GPT-5 nano (High) | OpenAI | Proprietary | – | 253 | 253 | – | 3,928 | |
| 121 | Gemini 1.5 Pro (001) | Google | Proprietary | – | 251 | 225 | 283 · 3 | 16,734 | |
| 122 | Llama 4 Maverick 17B 128E Instruct | Meta | Llama 4 | – | 251 | 251 | – | 6,923 | |
| 123 | Claude 3.5 Sonnet (20240620) | Anthropic | Proprietary | – | 250 | 250 | – | 21,624 | |
| 124 | Step-3 | StepFun | Apache 2.0 | – | 250 | 250 | – | 3,193 | |
| 125 | Qwen2.5-VL-72B-Instruct | Alibaba | Qwen | – | 250 | 227 | 278 · 16 | 3,768 | |
| 126 | Qwen2.5-VL-32B-Instruct | Alibaba | Apache 2.0 | – | 248 | 225 | 276 · 13 | 1,490 | |
| 127 | Mistral Small 3.2 (2506) | Mistral | Apache 2.0 | – | 245 | 245 | – | 10,843 | |
| 128 | Gemini 2.0 Flash-Lite Preview (02-05) | Google | Proprietary | – | 241 | 241 | – | 3,991 | |
| 129 | Pixtral Large (2411) | Mistral | MRL | – | 233 | 205 | 268 · 6 | 5,423 | |
| 130 | Llama 4 Scout 17B 16E Instruct | Meta | Llama | – | 233 | 233 | – | 6,446 | |
| 131 | Claude 3.5 Haiku (20241022) | Anthropic | Proprietary | – | 233 | 233 | – | 1,398 | |
| 132 | Mistral Small 3.1 24B Instruct (2503) | Mistral | Apache 2.0 | – | 233 | 233 | – | 28,490 | |
| 133 | Qwen2-VL-72B | Alibaba | Qwen | – | 231 | 197 | 276 · 8 | 5,937 | |
| 134 | Step-1o Vision 32K (High Res) | StepFun | Proprietary | – | 230 | 230 | – | 2,833 | |
| 135 | GPT-4.1 nano (2025-04-14) | OpenAI | Proprietary | $0.10 | 225 | 200 | 254 · 4 | 1,211 | |
| 136 | GPT-4o (2024-08-06) | OpenAI | Proprietary | $2.50 | 224 | 224 | – | 3,376 | |
| 137 | GPT-4o mini (2024-07-18) | OpenAI | Proprietary | – | 223 | 207 | 262 · 2 | 17,341 | |
| 138 | Gemini 1.5 Flash (001) | Google | Proprietary | – | 220 | 180 | 276 · 3 | 13,260 | |
| 139 | GPT-4 Turbo (2024-04-09) | OpenAI | Proprietary | $10.00 | 219 | 219 | – | 13,391 | |
| 139 | Molmo 2 8B | Ai2 | Apache 2.0 | – | 219 | 219 | – | 1,190 | |
| 140 | Gemini 1.5 Flash-8B (001) | Google | Proprietary | – | 218 | 188 | 256 · 3 | 6,243 | |
| 141 | Pixtral 12B (2409) | Mistral | Apache 2.0 | – | 201 | 159 | 259 · 5 | 7,511 | |
| 142 | Qwen-VL-Max (1119) | Alibaba | Proprietary | – | 197 | 197 | – | 1,422 | |
| 143 | Step-1V 32K | StepFun | Proprietary | – | 183 | 183 | – | 1,534 | |
| 144 | Claude 3 Opus (20240229) | Anthropic | Proprietary | – | 181 | 181 | – | 15,565 | |
| 145 | Molmo 72B (0924) | Ai2 | Apache 2.0 | – | 171 | 171 | – | 3,048 | |
| 146 | Hunyuan Standard Vision (2024-12-31) | Tencent | Proprietary | – | 169 | 169 | – | 809 | |
| 147 | Llama 3.2 90B Vision Instruct | Meta | Llama 3.2 | – | 163 | 163 | – | 8,682 | |
| 148 | Qwen2-VL-7B-Instruct | Alibaba | Apache 2.0 | – | 162 | 162 | – | 5,766 | |
| 149 | Phi-3.5 Vision Instruct |  Microsoft | MIT | – | 160 | 107 | 240 · 5 | 2,592 | |
| 150 | InternVL2 26B | Unknown | MIT | – | 159 | 159 | – | 5,148 | |
| 151 | Amazon Nova Lite 1.0 |  Amazon | Proprietary | – | 155 | 155 | – | 1,854 | |
| 152 | Amazon Nova Pro 1.0 | Amazon | Proprietary | – | 155 | 155 | – | 2,335 | |
| 153 | Claude 3 Sonnet (20240229) | Anthropic | Proprietary | – | 154 | 154 | – | 12,314 | |
| 154 | Yi Vision | Unknown | Proprietary | – | 146 | 146 | – | 1,219 | |
| 155 | Aya Vision 32B |  Cohere | CC-BY-NC-4.0 | – | 145 | 145 | – | 847 | |
| 156 | Claude 3 Haiku (20240307) | Anthropic | Proprietary | – | 145 | 145 | – | 13,380 | |
| 157 | Molmo 7B-D (0924) | Ai2 | Apache 2.0 | – | 142 | 142 | – | 2,815 | |
| 158 | Llama 3.2 11B Vision Instruct | Meta | Llama 3.2 | – | 141 | 141 | – | 4,817 | |
| 159 | nvila-internal-15b-v1 | NVIDIA | - | – | 137 | 137 | – | 1,077 | |
| 160 | LLaVA-OneVision Qwen2 72B OV | Unknown | Apache 2.0 | – | 134 | 134 | – | 1,321 | |
| 161 | LLaVA 1.6 34B | Unknown | Apache 2.0 | – | 127 | 127 | – | 4,531 | |
| 162 | CogVLM2 Llama3 Chat 19B | Z.ai | CogVLM2 | – | 127 | 127 | – | 1,991 | |
| 163 | MiniCPM-V 2.6 | Unknown | Apache 2.0 | – | 126 | 126 | – | 1,987 | |
| 164 | InternVL2 4B | Unknown | MIT | – | 124 | 124 | – | 3,703 | |
| 165 | Phi-3 Vision 128K Instruct | Microsoft | MIT | – | 91 | 91 | – | 1,401 |
Leaders by task
The top model for each kind of task, scored from votes on that task alone, and how many models were ranked.
| Task | Leader | Score | Models ranked |
|---|---|---|---|
| Captioning | Gemini 3.1 Pro Preview Google | 537 | 34 |
| Chinese | Gemini 3.8 Flash (High) Google | 815 | 128 |
| Creative writing about images | Gemini 3 Pro Google | 677 | 104 |
| Diagram | Claude Fable 5 (High) Anthropic | 750 | 123 |
| English | Claude Fable 5 (High) Anthropic | 608 | 167 |
| Homework | GPT-6.1 Sol (Max) OpenAI | 764 | 118 |
| Humor | Claude Fable 5 (High) Anthropic | 672 | 93 |
| Reading text in images | Claude Fable 5 (High) Anthropic | 680 | 123 |
| Recognising people and things | Gemini 3 Pro Google | 570 | 51 |
Related benchmarks
Fixed tests with right and wrong answers, scored out of 100%.
MMMU-Pro
College-level questions that need both an image and text to answer. · Top 10 of 74 · Higher is better
CharXiv-R
Reading and reasoning about charts from scientific papers. · Top 10 of 61 · Higher is better