Alibaba Cloud / Qwen Team model product
Qwen3-235B-A22B-Instruct-2507 is the updated instruct version of Qwen3-235B-A22B featuring significant improvements in general capabilities including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage. It provides substantial gains in long-tail knowledge coverage across multiple languages and markedly better alignment with user preferences in subjective and open-ended tasks.
Updated Aug 10, 2026. Default version: Qwen3-235B-A22B-Instruct-2507
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 Qwen3-235B-A22B-Instruct-2507.
| Creative Writing v3 | 0.9 | 1 | 13 | 100.0% | C | |
| MultiPL-E | 0.9 | 1 | 13 | 100.0% | C | |
| CSimpleQA | 0.8 | 2 | 7 | 83.3% | C | |
| ZebraLogic | 0.9 | 3 | 8 | 71.4% | C | |
| Arena-Hard v2 | 0.8 | 4 | 16 | 80.0% | C | |
| LiveBench 20241125 | 0.8 | 5 | 14 | 69.2% | C | |
| Multi-IF | 0.8 | 6 | 20 | 73.7% | C | |
| ARC-AGI | 0.4 | 7 | 7 | 0.0% | C | |
| WritingBench | 0.9 | 7 | 15 | 57.1% | C | |
| BFCL-v3 | 0.7 | 9 | 19 | 55.6% | C | |
| MMLU-ProX | 0.8 | 11 | 32 | 67.7% | C | |
| SimpleQA | 0.5 | 11 | 46 | 77.8% | C | |
| Include | 0.8 | 12 | 31 | 63.3% | C | |
| Aider-Polyglot | 0.6 | 14 | 22 | 38.1% | C | |
| MMLU-Redux | 0.9 | 14 | 48 | 72.3% | C | |
| PolyMATH | 0.5 | 14 | 23 | 40.9% | C | |
| SuperGPQA | 0.6 | 15 | 34 | 57.6% | C | |
| Tau2 Retail | 0.7 | 18 | 26 | 32.0% | C | |
| HMMT25 | 0.6 | 20 | 25 | 20.8% | C | |
| Tau2 Airline | 0.4 | 23 | 23 | 0.0% | C | |
| IFEval | 0.9 | 24 | 65 | 64.1% | C | |
| MMLU-Pro | 0.8 | 31 | 129 | 76.6% | C | |
| LiveCodeBench v6 | 0.5 | 45 | 53 | 15.4% | C | |
| AIME 2025 | 0.7 | 89 | 114 | 22.1% | C | |
| GPQA | 0.8 | 96 | 233 | 59.0% | C |
Preference and agent-evaluation signals from published Arena datasets.
| text factuality | korean | 19 | 1402.5 | 442 | N/A | |
| text factuality | german | 21 | 1418.1 | 539 | N/A | |
| text factuality | spanish | 35 | 1440.6 | 1,087 | N/A | |
| text factuality | industry mathematical | 42 | 1444.8 | 2,050 | N/A | |
| text factuality | polish | 44 | 1395.3 | 1,053 | N/A | |
| text factuality | chinese | 46 | 1486.9 | 2,000 | N/A | |
| text factuality | math | 50 | 1438.0 | 2,297 | N/A | |
| text factuality | expert | 53 | 1473.6 | 2,714 | N/A | |
| text factuality | french | 54 | 1450.1 | 991 | N/A | |
| text factuality | industry medicine and healthcare | 54 | 1470.6 | 2,366 | N/A | |
| text | german | 61 | 1435.6 | 1,707 | N/A | |
| text factuality | industry legal and government | 62 | 1455.4 | 2,639 | N/A | |
| text | korean | 63 | 1392.6 | 1,720 | N/A | |
| text factuality | industry life and physical and social science | 65 | 1463.0 | 6,421 | N/A | |
| text | french | 68 | 1449.3 | 1,980 | N/A | |
| text | expert | 69 | 1443.8 | 5,863 | N/A | |
| text | industry mathematical | 69 | 1439.9 | 4,902 | N/A | |
| text | japanese | 69 | 1388.9 | 1,265 | N/A | |
| text factuality | multi turn | 71 | 1446.9 | 7,374 | N/A | |
| text style control | french | 75 | 1458.6 | 1,980 | N/A | |
| text | industry medicine and healthcare | 76 | 1444.7 | 5,686 | N/A | |
| text style control | korean | 76 | 1384.0 | 1,720 | N/A | |
| text | longer query | 78 | 1427.2 | 25,604 | N/A | |
| text | multi turn | 78 | 1432.9 | 16,961 | N/A | |
| text | math | 79 | 1431.3 | 5,887 | N/A | |
| text factuality | industry business and management and financial operations | 79 | 1438.4 | 7,736 | N/A | |
| text | spanish | 80 | 1424.3 | 2,558 | N/A | |
| text factuality | longer query | 80 | 1447.5 | 14,022 | N/A | |
| text factuality | russian | 80 | 1428.8 | 3,825 | N/A | |
| text style control | japanese | 81 | 1381.8 | 1,265 | N/A | |
| text style control | industry medicine and healthcare | 82 | 1456.1 | 5,686 | N/A | |
| text | chinese | 83 | 1465.7 | 4,938 | N/A | |
| text factuality | instruction following | 83 | 1426.6 | 13,191 | N/A | |
| text | hard prompts | 84 | 1433.7 | 52,023 | N/A | |
| text factuality | non english | 85 | 1419.5 | 34,161 | N/A | |
| text style control | german | 85 | 1427.8 | 1,707 | N/A | |
| text | coding | 86 | 1444.7 | 21,199 | N/A | |
| text style control | spanish | 86 | 1422.8 | 2,558 | N/A | |
| text | industry software and it services | 87 | 1445.9 | 34,559 | N/A | |
| text | instruction following | 87 | 1408.5 | 26,840 | N/A | |
| text factuality | coding | 87 | 1478.3 | 10,147 | N/A | |
| text factuality | industry writing and literature and language | 87 | 1409.5 | 9,474 | N/A | |
| text | industry business and management and financial operations | 88 | 1417.7 | 17,875 | N/A | |
| text | industry life and physical and social science | 88 | 1438.4 | 15,251 | N/A | |
| text | non english | 88 | 1409.6 | 51,832 | N/A | |
| text factuality | hard prompts english | 88 | 1461.5 | 12,567 | N/A | |
| text style control | multi turn | 88 | 1437.5 | 16,961 | N/A | |
| text style control | industry business and management and financial operations | 89 | 1431.7 | 17,875 | N/A | |
| text factuality | hard prompts | 90 | 1453.7 | 36,451 | N/A | |
| text factuality | industry entertainment and sports and media | 90 | 1399.3 | 7,616 | N/A | |
| text style control | chinese | 90 | 1471.0 | 4,938 | N/A | |
| text | industry legal and government | 91 | 1429.6 | 6,484 | N/A | |
| text | hard prompts english | 92 | 1436.5 | 25,452 | N/A | |
| text factuality | overall | 92 | 1431.2 | 64,788 | N/A | |
| text factuality | creative writing | 92 | 1398.3 | 5,779 | N/A | |
| text style control | hard prompts | 92 | 1447.9 | 52,023 | N/A | |
| text style control | non english | 92 | 1411.3 | 51,832 | N/A | |
| text style control | expert | 93 | 1449.6 | 5,863 | N/A | |
| text factuality | industry software and it services | 94 | 1464.8 | 19,163 | N/A | |
| text | polish | 96 | 1407.0 | 3,810 | N/A | |
| text style control | industry software and it services | 96 | 1462.9 | 34,559 | N/A | |
| text style control | russian | 96 | 1416.6 | 8,378 | N/A | |
| text | russian | 97 | 1410.8 | 8,378 | N/A | |
| text factuality | exclude ties | 97 | 1425.8 | 45,944 | N/A | |
| text style control | industry legal and government | 97 | 1434.1 | 6,484 | N/A | |
| text style control | instruction following | 97 | 1415.8 | 26,840 | N/A | |
| text | exclude ties | 98 | 1407.6 | 68,887 | N/A | |
| text style control | coding | 98 | 1472.5 | 21,199 | N/A | |
| text style control | longer query | 98 | 1434.7 | 25,604 | N/A | |
| text factuality | english | 99 | 1438.6 | 28,958 | N/A | |
| text style control | industry life and physical and social science | 99 | 1445.7 | 15,251 | N/A | |
| text style control | industry mathematical | 99 | 1430.9 | 4,902 | N/A | |
| text | overall | 100 | 1419.2 | 96,721 | N/A | |
| text style control | hard prompts english | 102 | 1450.7 | 25,452 | N/A | |
| text style control | overall | 103 | 1422.9 | 96,721 | N/A | |
| text style control | exclude ties | 104 | 1414.6 | 68,887 | N/A | |
| text style control | math | 105 | 1418.2 | 5,887 | N/A | |
| text style control | polish | 105 | 1403.4 | 3,810 | N/A | |
| text | english | 108 | 1424.8 | 44,824 | N/A | |
| text | industry writing and literature and language | 109 | 1390.4 | 21,553 | N/A | |
| text style control | english | 111 | 1430.2 | 44,824 | N/A | |
| text | creative writing | 114 | 1375.8 | 13,473 | N/A | |
| text | industry entertainment and sports and media | 114 | 1374.6 | 17,608 | N/A | |
| text style control | industry writing and literature and language | 114 | 1396.8 | 21,553 | N/A | |
| text style control | creative writing | 124 | 1379.8 | 13,473 | N/A | |
| text style control | industry entertainment and sports and media | 124 | 1376.9 | 17,608 | 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.
| Cortecs | qwen3-235b-a22b-instruct-2507 | global | $0.069 | $0.455 | 262K | |
| LLM Gateway | qwen3-235b-a22b-instruct-2507 | global | $0.09 | $0.58 | 262.1K | |
| Venice AI | qwen3-235b-a22b-instruct-2507 | global | $0.15 | $0.75 | 128K | |
| 302.AI | qwen3-235b-a22b-instruct-2507 | global | $0.29 | $1.14 | 128K | |
| Scaleway | qwen3-235b-a22b-instruct-2507 | global | $0.75 | $2.25 | 260K | |
| GreenPT | qwen3-235b-a22b-instruct-2507 | global | $1.03 | $3.08 | 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.
| Qwen3-235B-A22B-Instruct-2507 | 49.0 | 235B | 262.1K | 16.4K | Yes | Apache 2.0 | |
| Qwen3 235B A22B | N/A | 235B | 131.1K | 16.4K | Yes | Apache 2.0 |
A concise description based on the published model registry.
Qwen3-235B-A22B-Instruct-2507 is the updated instruct version of Qwen3-235B-A22B featuring significant improvements in general capabilities including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage. It provides substantial gains in long-tail knowledge coverage across multiple languages and markedly better alignment with user preferences in subjective and open-ended 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 235B A22B.
Qwen3 235B A22B's default version was released on Jul 22, 2025.
No official standard PAYG price is currently available for Qwen3 235B A22B. The lowest tracked third-party offer starts at $0.069 input and $0.455 output via Cortecs.
Qwen3 235B A22B 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.
7 published provider offerings are linked to the default version.
Nearby ranked alternatives include Phi 4 Reasoning Plus, Nova Lite, Mistral Small 4.