# Arena vision / captioning (config: vision) > Source: Benchize (https://benchize.com) — https://benchize.com/benchmarks/benchmark%3Avision%3Acaptioning--config-vision. When citing or reusing this content, credit "Benchize" and link to https://benchize.com/benchmarks/benchmark%3Avision%3Acaptioning--config-vision. Human pairwise preferences on captioning prompts in Arena’s vision evaluation. The original organizer publishes model ratings and uncertainty; these are preference ratings rather than task accuracy. Canonical URL: https://benchize.com/benchmarks/benchmark%3Avision%3Acaptioning--config-vision Last updated: Wed, 21 Oct 2026 18:52:20 GMT Reproduction and quotation must name Benchize, link to the canonical HTML URL, and retain individual source citations. Human pairwise preferences on captioning prompts in Arena’s vision evaluation. The original organizer publishes model ratings and uncertainty; these are preference ratings rather than task accuracy. Multimodal · Models tested: 22 · Score publishers: [Arena](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset) ## Model results | # | Model | Score | | | --- | --- | --- | --- | | 1 | [Gemini 3.1 Pro Preview](https://benchize.com/models/gemini-3-1-pro-preview.md) | 1,344.89 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1344.8857647933394 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1296.3373370450227 rank 1 upper 1393.434192541656 variance 613.5559293611411 votes 137 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemini 3.1 Pro Preview Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/167/row` | | | | | 2 | [Gemini 3 Pro](https://benchize.com/models/gemini-3-pro.md) | 1,325.32 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1325.3215355192908 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1289.8742158507598 rank 2 upper 1360.768855187822 variance 327.09252665007 votes 218 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemini 3 Pro Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/168/row` | | | | | 3 | [Gemini 2.5 Pro](https://benchize.com/models/gemini-2-5-pro.md) | 1,303.57 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1303.5663600127425 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1280.9172187369518 rank 3 upper 1326.215501288533 variance 133.5387477713553 votes 796 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemini 2.5 Pro Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/169/row` | | | | | 4 | [Gemma 4 26B A4B](https://benchize.com/models/gemma-4-26b-a4b-it.md) | 1,296.71 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1296.7088279820423 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1226.9125957819335 rank 4 upper 1366.505060182151 variance 1268.1416765652693 votes 63 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemma 4 26B A4B Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/170/row` | | | | | 5 | [Qwen3 VL 235B A22B Instruct](https://benchize.com/models/qwen3-vl-235b-a22b-instruct.md) | 1,281.32 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1281.3240583191548 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1236.7939358957658 rank 5 upper 1325.854180742544 variance 516.1923882562198 votes 146 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Qwen3-VL-235B-A22B-Instruct Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/171/row` | | | | | 6 | [Gemini 3 Flash](https://benchize.com/models/gemini-3-flash.md) | 1,281.32 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1281.318117343302 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1242.5448803250006 rank 6 upper 1320.0913543616032 variance 391.35234270342113 votes 191 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemini 3 Flash Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/172/row` | | | | | 7 | [Gemini 2.5 Flash](https://benchize.com/models/gemini-2-5-flash.md) | 1,276.65 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1276.6460243375493 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1252.2293960521995 rank 7 upper 1301.062652622899 variance 155.19409803727018 votes 582 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemini 2.5 Flash Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/173/row` | | | | | 8 | [GPT-5.2](https://benchize.com/models/gpt-5-2-high.md)High | 1,266.64 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1266.6369794940363 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision Reasoning effort high thinkingEnabled true ### Reported statistics lower 1208.6705379679336 rank 8 upper 1324.6034210201387 variance 874.6959163268912 votes 94 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label GPT-5.2 (High) Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/174/row` | | | | | 9 | [Gemini 3.1 Flash Lite Preview](https://benchize.com/models/gemini-3-1-flash-lite-preview.md) | 1,259.45 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1259.4523996474286 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1194.9773876750426 rank 10 upper 1323.9274116198146 variance 1082.148049185171 votes 73 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemini 3.1 Flash-Lite Preview Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/176/row` | | | | | 10 | [GPT-5.1](https://benchize.com/models/gpt-5-1-high.md)High | 1,251.61 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1251.6128156563502 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision Reasoning effort high thinkingEnabled true ### Reported statistics lower 1196.2749679409571 rank 12 upper 1306.9506633717435 variance 797.1652262092222 votes 101 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label GPT-5.1 (High) Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/178/row` | | | | | 11 | [Gemini 3 Flash](https://benchize.com/models/gemini-3-flash-minimal.md)Minimal | 1,240.59 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1240.587267658662 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision Reasoning effort minimal thinkingEnabled true ### Reported statistics lower 1197.6733044677353 rank 14 upper 1283.501230849589 variance 479.4033524012782 votes 173 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemini 3 Flash (Minimal) Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/180/row` | | | | | 12 | [Qwen3.5 397B A17B](https://benchize.com/models/qwen3-5-397b-a17b.md) | 1,229.7 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1229.6993078212156 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1172.9633100356807 rank 15 upper 1286.4353056067505 variance 837.9559940560723 votes 78 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Qwen3.5-397B-A17B Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/181/row` | | | | | 13 | [Gemma 4 31B](https://benchize.com/models/gemma-4-31b-it.md) | 1,217.52 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1217.52490819928 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1163.4130633394807 rank 17 upper 1271.6367530590794 variance 762.2343205547891 votes 106 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemma 4 31B Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/183/row` | | | | | 14 | [GPT-5 Mini](https://benchize.com/models/gpt-5-mini-high.md)High | 1,216.96 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1216.962021342088 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision Reasoning effort high thinkingEnabled true ### Reported statistics lower 1180.4576750953474 rank 18 upper 1253.4663675888287 variance 346.89094875189556 votes 283 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label GPT-5 mini (High) Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/184/row` | | | | | 15 | [GPT-5 Chat](https://benchize.com/models/gpt-5-chat.md) | 1,211.57 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1211.5691736478948 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1179.8782197639878 rank 19 upper 1243.2601275318018 variance 261.44144840538655 votes 375 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label GPT-5 Chat Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/185/row` | | | | | 16 | [GPT-5](https://benchize.com/models/gpt-5-high.md)High | 1,203.76 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1203.7575443671406 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision Reasoning effort high thinkingEnabled true ### Reported statistics lower 1171.6813008398608 rank 23 upper 1235.8337878944205 variance 267.837154280741 votes 356 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label GPT-5 (High) Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/189/row` | | | | | 17 | [GPT-5.2](https://benchize.com/models/gpt-5-2.md) | 1,173.78 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1173.7762142675826 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1118.9092622014136 rank 26 upper 1228.6431663337514 variance 783.6560456705123 votes 107 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label GPT-5.2 Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/192/row` | | | | | 18 | [Qwen3.5-27B](https://benchize.com/models/qwen3-5-27b.md) | 1,170.2 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1170.204176393505 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1089.150812208778 rank 27 upper 1251.257540578232 variance 1710.1960875568866 votes 61 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Qwen3.5-27B Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/193/row` | | | | | 19 | [GPT-5.1](https://benchize.com/models/gpt-5-1.md) | 1,165.39 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1165.387625306252 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1116.4928335101135 rank 28 upper 1214.2824171023906 variance 622.341869684744 votes 123 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label GPT-5.1 Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/194/row` | | | | | 20 | [Mistral Small 3.1 24B](https://benchize.com/models/mistral-small-3-1-24b-instruct.md) | 1,152.89 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1152.8906231176825 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1114.4021706347094 rank 29 upper 1191.3790756006556 variance 385.6245878659224 votes 258 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Mistral Small 3.1 24B Instruct (2503) Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/195/row` | | | | | 21 | [Gemma 3 27B](https://benchize.com/models/gemma-3-27b-it.md) | 1,151.18 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1151.1786351856742 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1112.5530093767077 rank 30 upper 1189.8042609946406 variance 388.37822784853233 votes 258 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemma 3 27B IT Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/196/row` | | | | | 22 | [Gemini 2.0 Flash 001](https://benchize.com/models/gemini-2-0-flash-001.md) | 1,129.59 | | | ### Original source values \| Metric \| Value \| Unit \| Scope \| Direction \| \| --- \| --- \| --- \| --- \| --- \| \| `rating` \| 1129.5902452553232 \| rating \| reported \| Higher is better \| ### Configuration category captioning config vision ### Reported statistics lower 1068.683782295168 rank 32 upper 1190.4967082154785 variance 965.6740846297768 votes 106 ### Source [Original source](https://datasets-server.huggingface.co/parquet/lmarena-ai%2Fleaderboard-dataset/vision/latest) Source Arena Original model label Gemini 2.0 Flash (001) Observed 2026-10-21T18:45:05.190Z Evidence SHA-256 `06e61305486dce3fec5dc776ac2fd8321104e84c346cf9e08fa6f8e8a66ce0d7` Source locator `/rows/198/row` | | | | --- Cite as: Arena vision / captioning (config: vision) — Benchize, https://benchize.com/benchmarks/benchmark%3Avision%3Acaptioning--config-vision (retrieved 2026-10-21)