Alibaba Cloud / Qwen Team model product
Qwen3.5-397B-A17B is Qwen's flagship Mixture-of-Experts model with 397 billion total parameters and 17 billion activated parameters. It delivers state-of-the-art performance across knowledge, reasoning, coding, mathematics, multilingual understanding, instruction following, long context, and agent tasks.
Updated Aug 10, 2026. Default version: Qwen3.5-397B-A17B
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 Qwen3.5-397B-A17B.
| BrowseComp-zh | 0.7 | 1 | 13 | 100.0% | C | |
| NOVA-63 | 0.6 | 1 | 11 | 100.0% | C | |
| SecCodeBench | 0.7 | 1 | 1 | 100.0% | C | |
| VITA-Bench | 0.5 | 1 | 10 | 100.0% | C | |
| BFCL-V4 | 0.7 | 2 | 13 | 91.7% | C | |
| C-Eval | 0.9 | 2 | 18 | 94.1% | C | |
| LongBench v2 | 0.6 | 2 | 17 | 93.8% | C | |
| MMLU-Redux | 0.9 | 2 | 48 | 97.9% | C | |
| DeepPlanning | 0.3 | 3 | 9 | 75.0% | C | |
| HMMT25 | 0.9 | 3 | 25 | 91.7% | C | |
| Include | 0.9 | 3 | 31 | 93.3% | C | |
| MAXIFE | 0.9 | 3 | 11 | 80.0% | C | |
| MMLU-ProX | 0.8 | 3 | 32 | 93.5% | C | |
| AA-LCR | 0.7 | 4 | 15 | 78.6% | C | |
| PolyMATH | 0.7 | 4 | 23 | 86.4% | C | |
| Seal-0 | 0.5 | 4 | 6 | 40.0% | C | |
| t2-bench | 0.9 | 4 | 23 | 86.4% | C | |
| Global PIQA | 0.9 | 5 | 13 | 66.7% | C | |
| MMLU-Pro | 0.9 | 5 | 129 | 96.9% | C | |
| Multi-Challenge | 0.7 | 5 | 29 | 85.7% | C | |
| SuperGPQA | 0.7 | 5 | 34 | 87.9% | C | |
| WideSearch | 0.7 | 5 | 9 | 50.0% | C | |
| MCP-Mark | 0.5 | 6 | 8 | 28.6% | C | |
| HMMT 2025 | 0.9 | 7 | 33 | 81.3% | C | |
| IFBench | 0.8 | 8 | 28 | 74.1% | C | |
| IFEval | 0.9 | 8 | 65 | 89.1% | C | |
| WMT24++ | 0.8 | 8 | 23 | 68.2% | C | |
| AIME 2026 | 0.9 | 10 | 17 | 43.8% | C | |
| LiveCodeBench v6 | 0.8 | 12 | 53 | 78.8% | C | |
| IMO-AnswerBench | 0.8 | 14 | 19 | 27.8% | C | |
| MMMLU | 0.9 | 17 | 49 | 66.7% | C | |
| SWE-bench Multilingual | 0.7 | 20 | 34 | 42.4% | C | |
| Toolathlon | 0.4 | 26 | 31 | 16.7% | C | |
| BrowseComp | 0.7 | 29 | 58 | 50.9% | C | |
| GPQA | 0.9 | 29 | 233 | 87.9% | C | |
| SWE-Bench Verified | 0.8 | 29 | 104 | 72.8% | C | |
| Terminal-Bench 2.0 | 0.5 | 32 | 49 | 35.4% | C | |
| Humanity's Last Exam | 0.3 | 45 | 92 | 51.6% | C |
Preference and agent-evaluation signals from published Arena datasets.
| vision | entity recognition | 11 | 1277.8 | 180 | N/A | |
| text factuality | japanese | 13 | 1446.8 | 396 | N/A | |
| vision style control | entity recognition | 17 | 1226.3 | 180 | N/A | |
| text factuality | german | 23 | 1416.5 | 465 | N/A | |
| text factuality | korean | 25 | 1392.5 | 543 | N/A | |
| vision | creative writing vision | 25 | 1275.7 | 1,403 | N/A | |
| text factuality | industry mathematical | 26 | 1460.5 | 2,625 | N/A | |
| text | japanese | 28 | 1428.5 | 605 | N/A | |
| text factuality | chinese | 28 | 1506.1 | 2,590 | N/A | |
| text factuality | polish | 28 | 1427.7 | 776 | N/A | |
| vision | english | 28 | 1272.9 | 10,086 | N/A | |
| vision style control | creative writing vision | 28 | 1258.6 | 1,403 | N/A | |
| text factuality | math | 29 | 1456.9 | 2,622 | N/A | |
| vision | homework | 30 | 1287.8 | 3,064 | N/A | |
| vision | chinese | 31 | 1321.4 | 1,179 | N/A | |
| vision style control | humor | 31 | 1246.7 | 811 | N/A | |
| text factuality | french | 32 | 1476.9 | 1,552 | N/A | |
| vision | humor | 32 | 1265.1 | 811 | N/A | |
| vision style control | chinese | 32 | 1298.3 | 1,179 | N/A | |
| vision | overall | 34 | 1265.0 | 23,868 | N/A | |
| vision | diagram | 34 | 1280.3 | 6,340 | N/A | |
| vision style control | homework | 34 | 1282.5 | 3,064 | N/A | |
| vision | ocr | 35 | 1272.6 | 16,855 | N/A | |
| vision style control | english | 35 | 1254.0 | 10,086 | N/A | |
| vision style control | diagram | 39 | 1275.9 | 6,340 | N/A | |
| vision style control | ocr | 39 | 1261.9 | 16,855 | N/A | |
| text factuality | expert | 40 | 1483.8 | 5,002 | N/A | |
| text | chinese | 41 | 1500.8 | 3,394 | N/A | |
| text factuality | spanish | 41 | 1437.8 | 1,378 | N/A | |
| text | polish | 42 | 1445.2 | 1,275 | N/A | |
| text factuality | industry life and physical and social science | 42 | 1476.6 | 8,283 | N/A | |
| text factuality | industry medicine and healthcare | 42 | 1481.2 | 3,590 | N/A | |
| text style control | japanese | 42 | 1429.1 | 605 | N/A | |
| vision style control | overall | 42 | 1247.5 | 23,868 | N/A | |
| text | german | 43 | 1447.8 | 999 | N/A | |
| text style control | chinese | 44 | 1500.4 | 3,394 | N/A | |
| text | industry mathematical | 45 | 1458.9 | 3,500 | N/A | |
| text | french | 47 | 1460.1 | 2,097 | N/A | |
| text | industry life and physical and social science | 47 | 1458.4 | 10,304 | N/A | |
| text | math | 47 | 1451.5 | 3,622 | N/A | |
| text | spanish | 47 | 1450.0 | 1,950 | N/A | |
| text factuality | industry legal and government | 50 | 1462.4 | 3,595 | N/A | |
| text factuality | longer query | 50 | 1463.6 | 24,604 | N/A | |
| text style control | industry mathematical | 50 | 1460.3 | 3,500 | N/A | |
| text | expert | 51 | 1464.9 | 6,272 | N/A | |
| text | industry business and management and financial operations | 53 | 1432.5 | 12,916 | N/A | |
| text | industry medicine and healthcare | 53 | 1453.6 | 4,711 | N/A | |
| text | industry software and it services | 53 | 1463.2 | 25,731 | N/A | |
| text factuality | multi turn | 53 | 1460.3 | 8,680 | N/A | |
| text factuality | russian | 53 | 1452.7 | 5,018 | N/A | |
| text factuality | coding | 54 | 1498.2 | 16,334 | N/A | |
| text factuality | industry software and it services | 54 | 1483.8 | 24,660 | N/A | |
| text | longer query | 55 | 1442.9 | 26,491 | N/A | |
| text factuality | hard prompts english | 55 | 1478.3 | 18,313 | N/A | |
| text | non english | 56 | 1431.2 | 33,724 | N/A | |
| text | coding | 57 | 1465.3 | 18,329 | N/A | |
| text style control | expert | 58 | 1479.0 | 6,272 | N/A | |
| text style control | spanish | 58 | 1448.0 | 1,950 | N/A | |
| webdev | webdev-react | 58 | 1388.7 | 13,509 | N/A | |
| text | multi turn | 59 | 1445.7 | 10,769 | N/A | |
| text factuality | industry business and management and financial operations | 59 | 1449.8 | 10,692 | N/A | |
| text factuality | industry entertainment and sports and media | 59 | 1417.5 | 11,458 | N/A | |
| text factuality | instruction following | 59 | 1441.2 | 19,129 | N/A | |
| text style control | german | 59 | 1445.3 | 999 | N/A | |
| text style control | math | 59 | 1449.1 | 3,622 | N/A | |
| webdev | overall | 59 | 1400.0 | 16,511 | N/A | |
| webdev | webdev | 59 | 1400.0 | 16,511 | N/A | |
| text | exclude ties | 60 | 1434.5 | 46,583 | N/A | |
| text | hard prompts | 60 | 1448.5 | 41,548 | N/A | |
| text factuality | industry writing and literature and language | 60 | 1428.1 | 12,910 | N/A | |
| text factuality | non english | 60 | 1435.1 | 33,524 | N/A | |
| text | overall | 61 | 1438.5 | 63,719 | N/A | |
| text | instruction following | 61 | 1422.5 | 20,914 | N/A | |
| text factuality | hard prompts | 61 | 1468.1 | 41,317 | N/A | |
| webdev | webdev-html | 61 | 1412.2 | 2,017 | N/A | |
| text factuality | overall | 62 | 1445.4 | 63,388 | N/A | |
| text | hard prompts english | 63 | 1451.4 | 20,256 | N/A | |
| text | russian | 63 | 1431.4 | 6,539 | N/A | |
| text factuality | exclude ties | 63 | 1445.8 | 46,350 | N/A | |
| text style control | french | 64 | 1467.1 | 2,097 | N/A | |
| text | industry writing and literature and language | 65 | 1416.1 | 14,873 | N/A | |
| text style control | longer query | 65 | 1453.9 | 26,491 | N/A | |
| text | english | 67 | 1443.8 | 29,993 | N/A | |
| text factuality | english | 67 | 1453.4 | 29,737 | N/A | |
| text style control | industry life and physical and social science | 67 | 1465.8 | 10,304 | N/A | |
| text style control | non english | 67 | 1433.3 | 33,724 | N/A | |
| text style control | polish | 68 | 1441.9 | 1,275 | N/A | |
| text | korean | 69 | 1386.8 | 1,046 | N/A | |
| text factuality | creative writing | 69 | 1414.2 | 8,165 | N/A | |
| text style control | coding | 69 | 1491.7 | 18,329 | N/A | |
| text style control | multi turn | 69 | 1452.2 | 10,769 | N/A | |
| text style control | hard prompts | 70 | 1463.4 | 41,548 | N/A | |
| text style control | hard prompts english | 72 | 1467.4 | 20,256 | N/A | |
| text style control | industry software and it services | 72 | 1480.5 | 25,731 | N/A | |
| text style control | instruction following | 72 | 1432.7 | 20,914 | N/A | |
| text | industry legal and government | 73 | 1438.0 | 4,743 | N/A | |
| text style control | overall | 73 | 1442.2 | 63,719 | N/A | |
| text style control | industry medicine and healthcare | 73 | 1462.6 | 4,711 | N/A | |
| text style control | industry writing and literature and language | 73 | 1421.5 | 14,873 | N/A | |
| text style control | exclude ties | 74 | 1441.3 | 46,583 | 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.
| DigitalOcean | qwen3.5-397b-a17b | global | $0.3025 | $1.93 | 131.1K | |
| CrofAI | qwen3.5-397b-a17b | global | $0.35 | $1.75 | 262.1K | |
| Alibaba (China) | qwen3.5-397b-a17b | global | $0.43 | $2.58 | 262.1K | |
| Scaleway | qwen3.5-397b-a17b | global | $0.60 | $3.6 | 256K | |
| Cortecs | qwen3.5-397b-a17b | global | $0.668 | $4.01 | 262K | |
| OVHcloud AI Endpoints | qwen3.5-397b-a17b | global | $0.71 | $4.25 | 262.1K | |
| GreenPT | qwen3.5-397b-a17b | global | $0.798 | $4.96 | 262.1K |
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.
| Qwen3.5-397B-A17B | 75.2 | 397B | 262.1K | 65.5K | Yes | Apache 2.0 |
A concise description based on the published model registry.
Qwen3.5-397B-A17B is Qwen's flagship Mixture-of-Experts model with 397 billion total parameters and 17 billion activated parameters. It delivers state-of-the-art performance across knowledge, reasoning, coding, mathematics, multilingual understanding, instruction following, long context, and agent tasks.
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 397B A17B.
Qwen3.5 397B A17B's default version was released on Feb 16, 2026.
Qwen3.5 397B A17B's official API price is $0.43 per million input tokens and $2.58 per million output tokens via Alibaba (China). The lowest tracked third-party offer starts at $0.3025 input and $1.93 output via DigitalOcean.
Qwen3.5 397B A17B 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.
8 published provider offerings are linked to the default version.
Nearby ranked alternatives include GPT-5.5, GPT-5.4, DeepSeek-V3.2.