Alibaba Cloud / Qwen Team model product
Qwen3.6-27B is a dense 27-billion-parameter multimodal model in the Qwen3.6 series, supporting both vision-language thinking and non-thinking modes in a single unified checkpoint. The 64-layer language model uses a hybrid layout of 16 repeats of (3 × Gated DeltaNet → FFN, 1 × Gated Attention → FFN) with hidden dim 5120 and FFN intermediate 17408 — Gated DeltaNet has 48/16 heads for V/QK (head dim 128) and Gated Attention has 24/4 heads for Q/KV (head dim 256). It supports a native 262,144-token context extensible to ~1,010,000 via YaRN and is trained with multi-token prediction. The release delivers flagship-level agentic coding, surpassing the previous-generation open-source flagship Qwen3.5-397B-A17B (397B total / 17B active) on every major coding benchmark including SWE-bench Verified (77.2), SWE-bench Pro (53.5), Terminal-Bench 2.0 (59.3), and SkillsBench (48.2), and reaches 87.8 on GPQA Diamond. Released as open weights under Apache 2.0; accessible via Qwen Studio with the Alibaba Cloud Model Studio API coming soon.
Updated Aug 10, 2026. Default version: Qwen3.6-27B
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 Qwen3.6-27B.
| EmbSpatialBench | 0.8 | 1 | 8 | 100.0% | C | |
| RefSpatialBench | 0.7 | 1 | 6 | 100.0% | C | |
| MVBench | 0.8 | 2 | 17 | 93.8% | C | |
| QwenWebBench | 1487.0 | 2 | 2 | 0.0% | C | |
| RefCOCO-avg | 0.9 | 2 | 7 | 83.3% | C | |
| VideoMME w sub. | 0.9 | 2 | 10 | 88.9% | C | |
| VLMsAreBlind | 1.0 | 2 | 4 | 66.7% | C | |
| CountBench | 1.0 | 3 | 6 | 60.0% | C | |
| V* | 0.9 | 3 | 7 | 66.7% | C | |
| ZClawBench | 0.5 | 3 | 4 | 33.3% | C | |
| AndroidWorld | 0.7 | 4 | 4 | 0.0% | C | |
| DynaMath | 0.9 | 4 | 7 | 50.0% | C | |
| HMMT25 | 0.9 | 4 | 25 | 87.5% | C | |
| MathVista-Mini | 0.9 | 4 | 23 | 86.4% | C | |
| MLVU | 0.9 | 4 | 10 | 66.7% | C | |
| MMBench-V1.1 | 0.9 | 4 | 18 | 82.3% | C | |
| MMStar | 0.8 | 4 | 22 | 85.7% | C | |
| SkillsBench | 0.5 | 5 | 7 | 33.3% | C | |
| CC-OCR | 0.8 | 6 | 18 | 70.6% | C | |
| C-Eval | 0.9 | 6 | 18 | 70.6% | C | |
| AIME 2026 | 0.9 | 7 | 17 | 62.5% | C | |
| MMMU | 0.8 | 8 | 63 | 88.7% | C | |
| SuperGPQA | 0.7 | 8 | 34 | 78.8% | C | |
| HMMT Feb 26 | 0.8 | 9 | 11 | 20.0% | C | |
| MMLU-Redux | 0.9 | 9 | 48 | 83.0% | C | |
| OCRBench | 0.9 | 9 | 22 | 61.9% | C | |
| RealWorldQA | 0.8 | 9 | 26 | 68.0% | C | |
| VideoMMMU | 0.8 | 9 | 26 | 68.0% | C | |
| Claw-Eval | 0.6 | 10 | 13 | 25.0% | C | |
| ERQA | 0.6 | 10 | 23 | 59.1% | C | |
| HMMT 2025 | 0.9 | 11 | 33 | 68.8% | C | |
| LiveCodeBench v6 | 0.8 | 11 | 53 | 80.8% | C | |
| MMLU-Pro | 0.9 | 12 | 129 | 91.4% | C | |
| SimpleVQA | 0.6 | 12 | 13 | 8.3% | C | |
| NL2Repo | 0.4 | 13 | 14 | 7.7% | C | |
| IMO-AnswerBench | 0.8 | 15 | 19 | 22.2% | C | |
| SWE-bench Multilingual | 0.7 | 17 | 34 | 51.5% | C | |
| Terminal-Bench 2.0 | 0.6 | 22 | 49 | 56.3% | C | |
| CharXiv-R | 0.8 | 23 | 47 | 52.2% | C | |
| SWE-Bench Verified | 0.8 | 26 | 104 | 75.7% | C | |
| MMMU-Pro | 0.8 | 30 | 65 | 54.7% | C | |
| SWE-Bench Pro | 0.5 | 36 | 44 | 18.6% | C | |
| GPQA | 0.9 | 37 | 233 | 84.5% | C | |
| Humanity's Last Exam | 0.2 | 52 | 92 | 44.0% | C |
Preference and agent-evaluation signals from published Arena datasets.
The default version has no published Arena rows, or its source alias has not been resolved.
Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.
| QVAC | qwen3.6-27b | global | N/A | N/A | 262.1K | |
| CrofAI | qwen3.6-27b | global | $0.20 | $1.5 | 262.1K | |
| Cortecs | qwen3.6-27b | global | $0.446 | $3.01 | 262K | |
| OVHcloud AI Endpoints | qwen3.6-27b | global | $0.47 | $3.19 | 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.
| Qwen3.6-27B | 72.2 | 27.8B | 262.1K | 65.5K | Yes | Apache 2.0 |
A concise description based on the published model registry.
Qwen3.6-27B is a dense 27-billion-parameter multimodal model in the Qwen3.6 series, supporting both vision-language thinking and non-thinking modes in a single unified checkpoint. The 64-layer language model uses a hybrid layout of 16 repeats of (3 × Gated DeltaNet → FFN, 1 × Gated Attention → FFN) with hidden dim 5120 and FFN intermediate 17408 — Gated DeltaNet has 48/16 heads for V/QK (head dim 128) and Gated Attention has 24/4 heads for Q/KV (head dim 256). It supports a native 262,144-token context extensible to ~1,010,000 via YaRN and is trained with multi-token prediction. The release delivers flagship-level agentic coding, surpassing the previous-generation open-source flagship Qwen3.5-397B-A17B (397B total / 17B active) on every major coding benchmark including SWE-bench Verified (77.2), SWE-bench Pro (53.5), Terminal-Bench 2.0 (59.3), and SkillsBench (48.2), and reaches 87.8 on GPQA Diamond. Released as open weights under Apache 2.0; accessible via Qwen Studio with the Alibaba Cloud Model Studio API coming soon.
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 Qwen3.6 27B.
Qwen3.6 27B's default version was released on Apr 21, 2026.
No official standard PAYG price is currently available for Qwen3.6 27B. The lowest tracked third-party offer starts at $0.20 input and $1.5 output via CrofAI.
Qwen3.6 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.
5 published provider offerings are linked to the default version.
Nearby ranked alternatives include Qwen3.7 Plus, Gemini 3 Flash, Qwen3.6 Plus.