Qwen3-VL-235B-A22B Instruct is an open-weight multimodal model that unifies strong text generation with visual understanding across images and video. The Instruct model targets general vision-language use (VQA, document parsing, chart/table...
text+image->text| Provider | Region / Tag | Context | Input/M | Output/M | Uptime 1d | Latency | Quant | Cache |
|---|---|---|---|---|---|---|---|---|
| Alibaba | alibaba | 131K | $0.260 | $1.04 | 99.97% | - | - | - |
| Parasail | parasail/fp8 | 131K | $0.210 | $1.90 | 97.91% | - | fp8 | - |
| DeepInfra | deepinfra/fp8 | 262K | $0.200 | $0.880 | 96.43% | - | fp8 | - |
| Venice | venice/fp8 | 256K | $0.250 | $1.50 | 93.42% | - | fp8 | - |
| Novita | novita/bf16 | 131K | $0.300 | $1.50 | 89.82% | - | bf16 | - |
| AtlasCloud | atlas-cloud/fp8 | 131K | $0.300 | $1.50 | 86.92% | - | fp8 | - |
| Rank | Category | Variant | Trend | Elo | Votes |
|---|---|---|---|---|---|
| #65 | French | - | 1455 | 234 | |
| #66 | Industry Business And Management And Financial Operations | - | 1439 | 2,233 | |
| #67 | Korean | - | 1382 | 308 | |
| #76 | Expert | - | 1449 | 563 | |
| #79 | Industry Legal And Government | - | 1438 | 754 | |
| #80 | Japanese | - | 1369 | 151 | |
| #83 | Hard Prompts English | - | 1452 | 2,953 | |
| #84 | Industry Life And Physical And Social Science | - | 1440 | 1,751 | |
| #85 | German | - | 1410 | 259 | |
| #85 | Multi Turn | - | 1426 | 2,014 | |
| #86 | Coding | - | 1466 | 2,316 | |
| #86 | Industry Software And It Services | - | 1455 | 4,096 | |
| #87 | Instruction Following | - | 1414 | 2,925 | |
| #88 | Chinese | - | 1455 | 494 | |
| #89 | Hard Prompts | - | 1440 | 5,717 | |
| #94 | Longer Query | - | 1425 | 2,204 | |
| #97 | Industry Medicine And Healthcare | - | 1436 | 628 | |
| #98 | English | - | 1427 | 5,647 | |
| #99 | Exclude Ties | - | 1402 | 8,127 | |
| #99 | Spanish | - | 1400 | 406 | |
| #101 | Math | - | 1411 | 704 | |
| #101 | Overall | - | 1415 | 11,512 | |
| #106 | Russian | - | 1401 | 717 | |
| #108 | Industry Writing And Literature And Language | - | 1387 | 2,364 | |
| #114 | Industry Mathematical | - | 1411 | 587 | |
| #123 | Industry Entertainment And Sports And Media | - | 1361 | 1,906 | |
| #126 | Creative Writing | - | 1362 | 1,382 |
| Rank | Category | Variant | Trend | Elo | Votes |
|---|---|---|---|---|---|
| #12 | Captioning | - | 1207 | 148 | |
| #22 | Entity Recognition | - | 1194 | 148 | |
| #38 | Homework | - | 1260 | 1,612 | |
| #41 | Creative Writing Vision | - | 1231 | 819 | |
| #48 | Chinese | - | 1245 | 721 | |
| #48 | English | - | 1218 | 5,123 | |
| #48 | Ocr | - | 1230 | 7,597 | |
| #50 | Diagram | - | 1241 | 2,687 | |
| #50 | Overall | - | 1215 | 12,202 | |
| #51 | Creative Writing | - | 1200 | 819 | |
| #53 | Humor | - | 1160 | 454 |