Alibaba Cloud / Qwen Team model product
Qwen3.5-27B is a multimodal dense foundation model with 27 billion parameters. It combines strong reasoning, coding, multilingual, long-context, and visual understanding performance in a production-friendly open-weight package with a native 262K context window.
Updated Aug 10, 2026. Default version: Qwen3.5-27B
Structured fields from the published default version.
This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.
Published benchmark records for the scored version Qwen3.5-27B.
| CountBench | 1.0 | 1 | 6 | 100.0% | C | |
| Hallusion Bench | 0.7 | 1 | 16 | 100.0% | C | |
| IFEval | 0.9 | 1 | 65 | 100.0% | C | |
| DynaMath | 0.9 | 2 | 7 | 83.3% | C | |
| EmbSpatialBench | 0.8 | 2 | 8 | 85.7% | C | |
| FullStackBench en | 0.6 | 2 | 3 | 50.0% | C | |
| FullStackBench zh | 0.6 | 2 | 3 | 50.0% | C | |
| Hypersim | 0.1 | 2 | 4 | 66.7% | C | |
| LingoQA | 0.8 | 2 | 4 | 66.7% | C | |
| MathVista-Mini | 0.9 | 2 | 23 | 95.5% | C | |
| MMLongBench-Doc | 0.6 | 2 | 5 | 75.0% | C | |
| Nuscene | 0.2 | 2 | 3 | 50.0% | C | |
| PMC-VQA | 0.6 | 2 | 3 | 50.0% | C | |
| SlakeVQA | 0.8 | 2 | 4 | 66.7% | C | |
| SUNRGBD | 0.4 | 2 | 4 | 66.7% | C | |
| TIR-Bench | 0.6 | 2 | 4 | 66.7% | C | |
| VideoMME w/o sub. | 0.8 | 2 | 10 | 88.9% | C | |
| ZEROBench-Sub | 0.4 | 2 | 5 | 75.0% | C | |
| AndroidWorld_SR | 0.6 | 3 | 8 | 71.4% | C | |
| MedXpertQA | 0.6 | 3 | 12 | 81.8% | C | |
| MMBench-V1.1 | 0.9 | 3 | 18 | 88.2% | C | |
| MMVU | 0.7 | 3 | 4 | 33.3% | C | |
| OJBench | 0.4 | 3 | 9 | 75.0% | C | |
| Seal-0 | 0.5 | 3 | 6 | 60.0% | C | |
| VLMsAreBlind | 1.0 | 3 | 4 | 33.3% | C | |
| MVBench | 0.7 | 4 | 17 | 81.3% | C | |
| RefSpatialBench | 0.7 | 4 | 6 | 40.0% | C | |
| V* | 0.9 | 4 | 7 | 50.0% | C | |
| VideoMME w sub. | 0.9 | 4 | 10 | 66.7% | C | |
| BFCL-V4 | 0.7 | 5 | 13 | 66.7% | C | |
| MAXIFE | 0.9 | 5 | 11 | 60.0% | C | |
| MMStar | 0.8 | 5 | 22 | 81.0% | C | |
| NOVA-63 | 0.6 | 5 | 11 | 60.0% | C | |
| PolyMATH | 0.7 | 5 | 23 | 81.8% | C | |
| VITA-Bench | 0.4 | 5 | 10 | 55.6% | C | |
| MLVU | 0.9 | 6 | 10 | 44.4% | C | |
| RefCOCO-avg | 0.9 | 6 | 7 | 16.7% | C | |
| WideSearch | 0.6 | 6 | 9 | 37.5% | C | |
| ZEROBench | 0.1 | 6 | 9 | 37.5% | C | |
| AA-LCR | 0.7 | 7 | 15 | 57.1% | C | |
| AI2D | 0.9 | 7 | 32 | 80.7% | C | |
| BabyVision | 0.4 | 7 | 9 | 25.0% | C | |
| CC-OCR | 0.8 | 7 | 18 | 64.7% | C | |
| C-Eval | 0.9 | 7 | 18 | 64.7% | C | |
| CodeForces | 0.8 | 7 | 16 | 60.0% | C | |
| DeepPlanning | 0.2 | 7 | 9 | 25.0% | C | |
| HMMT25 | 0.9 | 7 | 25 | 75.0% | C | |
| IFBench | 0.8 | 7 | 28 | 77.8% | C | |
| Include | 0.8 | 7 | 31 | 80.0% | C | |
| LVBench | 0.7 | 7 | 24 | 73.9% | C | |
| MMLU-ProX | 0.8 | 7 | 32 | 80.7% | C | |
| Global PIQA | 0.9 | 8 | 13 | 41.7% | C | |
| MathVision | 0.9 | 8 | 32 | 77.4% | C | |
| OCRBench | 0.9 | 8 | 22 | 66.7% | C | |
| OmniDocBench 1.5 | 0.9 | 8 | 16 | 53.3% | C | |
| ScreenSpot Pro | 0.7 | 8 | 24 | 69.6% | C | |
| BrowseComp-zh | 0.6 | 9 | 13 | 33.3% | C | |
| MMMU | 0.8 | 9 | 63 | 87.1% | C | |
| Multi-Challenge | 0.6 | 9 | 29 | 71.4% | C | |
| SuperGPQA | 0.7 | 9 | 34 | 75.8% | C | |
| LongBench v2 | 0.6 | 10 | 17 | 43.8% | C | |
| RealWorldQA | 0.8 | 10 | 26 | 64.0% | C | |
| WMT24++ | 0.8 | 10 | 23 | 59.1% | C | |
| ERQA | 0.6 | 12 | 23 | 50.0% | C | |
| MMLU-Pro | 0.9 | 13 | 129 | 90.6% | C | |
| MMLU-Redux | 0.9 | 13 | 48 | 74.5% | C | |
| SimpleVQA | 0.6 | 13 | 13 | 0.0% | C | |
| HMMT 2025 | 0.9 | 14 | 33 | 59.4% | C | |
| ODinW | 0.4 | 14 | 16 | 13.3% | C | |
| VideoMMMU | 0.8 | 14 | 26 | 48.0% | C | |
| t2-bench | 0.8 | 15 | 23 | 36.4% | C | |
| LiveCodeBench v6 | 0.8 | 18 | 53 | 67.3% | C | |
| OSWorld-Verified | 0.6 | 20 | 22 | 9.5% | C | |
| CharXiv-R | 0.8 | 21 | 47 | 56.5% | C | |
| Humanity's Last Exam | 0.5 | 22 | 92 | 76.9% | C | |
| MMMLU | 0.9 | 26 | 49 | 47.9% | C | |
| BrowseComp | 0.6 | 34 | 58 | 42.1% | C | |
| MMMU-Pro | 0.8 | 34 | 65 | 48.4% | C | |
| Terminal-Bench 2.0 | 0.4 | 43 | 49 | 12.5% | C | |
| SWE-Bench Verified | 0.7 | 51 | 104 | 51.5% | C | |
| GPQA | 0.9 | 52 | 233 | 78.0% | C |
Preference and agent-evaluation signals from published Arena datasets.
| vision | entity recognition | 32 | 1206.2 | 146 | N/A | |
| vision | chinese | 35 | 1308.4 | 1,150 | N/A | |
| vision style control | chinese | 41 | 1279.0 | 1,150 | N/A | |
| vision style control | entity recognition | 41 | 1149.2 | 146 | N/A | |
| text factuality | industry mathematical | 47 | 1439.3 | 1,024 | N/A | |
| text factuality | math | 47 | 1440.6 | 1,066 | N/A | |
| vision | humor | 47 | 1230.5 | 736 | N/A | |
| vision style control | humor | 51 | 1202.8 | 736 | N/A | |
| vision | english | 52 | 1245.4 | 9,069 | N/A | |
| vision | homework | 53 | 1253.0 | 2,866 | N/A | |
| vision | overall | 55 | 1240.9 | 21,450 | N/A | |
| vision | ocr | 55 | 1247.8 | 15,381 | N/A | |
| text factuality | chinese | 56 | 1477.1 | 1,013 | N/A | |
| vision | creative writing vision | 57 | 1223.4 | 1,201 | N/A | |
| vision | diagram | 57 | 1249.6 | 5,704 | N/A | |
| vision style control | english | 58 | 1223.1 | 9,069 | N/A | |
| vision style control | ocr | 58 | 1232.3 | 15,381 | N/A | |
| vision style control | overall | 59 | 1218.7 | 21,450 | N/A | |
| vision style control | creative writing vision | 61 | 1198.4 | 1,201 | N/A | |
| vision style control | diagram | 61 | 1241.8 | 5,704 | N/A | |
| vision style control | homework | 61 | 1241.7 | 2,866 | N/A | |
| text | chinese | 65 | 1475.8 | 1,504 | N/A | |
| text factuality | industry medicine and healthcare | 67 | 1464.3 | 1,283 | N/A | |
| webdev | webdev-html | 68 | 1392.9 | 867 | N/A | |
| webdev | webdev-react | 75 | 1345.5 | 6,351 | N/A | |
| text | math | 77 | 1432.5 | 1,634 | N/A | |
| webdev | overall | 77 | 1357.5 | 7,303 | N/A | |
| webdev | webdev | 77 | 1357.5 | 7,303 | N/A | |
| text | industry mathematical | 79 | 1434.2 | 1,506 | N/A | |
| text factuality | industry life and physical and social science | 83 | 1455.4 | 3,024 | N/A | |
| text factuality | expert | 85 | 1442.8 | 1,632 | N/A | |
| text factuality | russian | 87 | 1413.8 | 1,861 | N/A | |
| text style control | math | 89 | 1427.6 | 1,634 | N/A | |
| text style control | chinese | 91 | 1470.2 | 1,504 | N/A | |
| text style control | industry mathematical | 93 | 1434.7 | 1,506 | N/A | |
| text | industry life and physical and social science | 94 | 1434.7 | 4,435 | N/A | |
| text factuality | industry legal and government | 94 | 1428.5 | 1,415 | N/A | |
| text factuality | hard prompts english | 95 | 1455.6 | 6,758 | N/A | |
| text factuality | instruction following | 96 | 1414.6 | 6,799 | N/A | |
| text factuality | multi turn | 96 | 1425.2 | 3,182 | N/A | |
| text | expert | 97 | 1425.4 | 2,354 | N/A | |
| text factuality | longer query | 97 | 1433.6 | 8,468 | N/A | |
| text | japanese | 98 | 1354.8 | 201 | N/A | |
| text | spanish | 98 | 1412.6 | 837 | N/A | |
| text factuality | industry business and management and financial operations | 99 | 1425.5 | 3,657 | N/A | |
| text | longer query | 100 | 1413.2 | 10,115 | N/A | |
| text factuality | coding | 101 | 1465.3 | 5,730 | N/A | |
| text | german | 103 | 1400.1 | 441 | N/A | |
| text factuality | english | 103 | 1434.3 | 12,574 | N/A | |
| text style control | industry life and physical and social science | 103 | 1439.4 | 4,435 | N/A | |
| text factuality | industry writing and literature and language | 104 | 1396.8 | 4,353 | N/A | |
| text factuality | industry software and it services | 105 | 1456.5 | 9,726 | N/A | |
| text factuality | creative writing | 108 | 1375.5 | 2,798 | N/A | |
| text | hard prompts english | 109 | 1425.3 | 8,370 | N/A | |
| text factuality | overall | 109 | 1419.0 | 26,873 | N/A | |
| text style control | japanese | 109 | 1348.6 | 201 | N/A | |
| text style control | spanish | 110 | 1404.2 | 837 | N/A | |
| text | english | 111 | 1422.7 | 12,752 | N/A | |
| text | industry medicine and healthcare | 111 | 1422.6 | 1,909 | N/A | |
| text factuality | exclude ties | 111 | 1409.3 | 19,645 | N/A | |
| text factuality | industry entertainment and sports and media | 111 | 1376.8 | 3,616 | N/A | |
| text style control | expert | 111 | 1438.0 | 2,354 | N/A | |
| text | korean | 112 | 1354.2 | 439 | N/A | |
| text | industry software and it services | 115 | 1430.2 | 10,657 | N/A | |
| text | multi turn | 115 | 1409.8 | 4,553 | N/A | |
| text style control | german | 115 | 1397.1 | 441 | N/A | |
| text | industry business and management and financial operations | 116 | 1399.1 | 5,298 | N/A | |
| text | instruction following | 116 | 1392.6 | 8,492 | N/A | |
| text style control | hard prompts english | 116 | 1440.7 | 8,370 | N/A | |
| text style control | longer query | 116 | 1422.2 | 10,115 | N/A | |
| text factuality | hard prompts | 117 | 1438.5 | 16,756 | N/A | |
| text factuality | non english | 117 | 1399.7 | 14,174 | N/A | |
| text | coding | 118 | 1425.2 | 7,446 | N/A | |
| text style control | english | 118 | 1425.8 | 12,752 | N/A | |
| text style control | multi turn | 119 | 1413.8 | 4,553 | N/A | |
| text | overall | 122 | 1408.0 | 27,052 | N/A | |
| text | exclude ties | 122 | 1390.1 | 19,782 | N/A | |
| text | hard prompts | 122 | 1413.7 | 16,893 | N/A | |
| text style control | industry business and management and financial operations | 122 | 1408.2 | 5,298 | N/A | |
| text style control | industry software and it services | 122 | 1445.2 | 10,657 | N/A | |
| text style control | korean | 122 | 1346.7 | 439 | N/A | |
| text style control | instruction following | 123 | 1400.6 | 8,492 | N/A | |
| text | french | 124 | 1405.1 | 787 | N/A | |
| text | polish | 125 | 1378.4 | 519 | N/A | |
| text | russian | 125 | 1390.1 | 2,812 | N/A | |
| text style control | hard prompts | 126 | 1426.9 | 16,893 | N/A | |
| text | industry writing and literature and language | 127 | 1379.8 | 6,086 | N/A | |
| text | industry legal and government | 128 | 1406.6 | 2,069 | N/A | |
| text | non english | 128 | 1388.8 | 14,300 | N/A | |
| text style control | overall | 129 | 1407.9 | 27,052 | N/A | |
| text style control | exclude ties | 129 | 1391.4 | 19,782 | N/A | |
| text style control | coding | 130 | 1450.0 | 7,446 | N/A | |
| text | creative writing | 131 | 1361.2 | 4,019 | N/A | |
| text style control | industry medicine and healthcare | 131 | 1428.2 | 1,909 | N/A | |
| text | industry entertainment and sports and media | 132 | 1359.7 | 5,012 | N/A | |
| text style control | french | 132 | 1411.2 | 787 | N/A | |
| text style control | industry legal and government | 134 | 1406.6 | 2,069 | N/A | |
| text style control | industry writing and literature and language | 136 | 1379.8 | 6,086 | N/A | |
| text style control | non english | 136 | 1386.0 | 14,300 | N/A | |
| text style control | russian | 137 | 1390.3 | 2,812 | N/A |
Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.
| NanoGPT | qwen3.5-27b | global | $0.27 | $2.16 | 260.1K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.
| Qwen3.5-27B | 57.1 | 27B | 262.1K | 65.5K | Yes | Apache 2.0 |
A concise description based on the published model registry.
Qwen3.5-27B is a multimodal dense foundation model with 27 billion parameters. It combines strong reasoning, coding, multilingual, long-context, and visual understanding performance in a production-friendly open-weight package with a native 262K context window.
Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.
Data snapshot: 2026-08-07. Editorial model content is not available in the backend.
Open a comparison with the three ranked models immediately above and below this model.
Recommendations prioritize the same model type and family, then the closest published LLMBoard score.
Common questions about Qwen3.5 27B.
Qwen3.5 27B's default version was released on Feb 24, 2026.
No official standard PAYG price is currently available for Qwen3.5 27B. The lowest tracked third-party offer starts at $0.27 input and $2.16 output via NanoGPT.
Qwen3.5 27B is published under Alibaba Cloud / Qwen Team in the model registry.
The default version has a 262.1K token context window.
Yes. The default version is marked as open weight under Apache 2.0.
1 published provider offerings are linked to the default version.
Nearby ranked alternatives include Claude Opus 3, GPT-4.5, Qwen3 VL 235B A22B Thinking.