Model Details
Qwen3.6 Plus
Quick answer: Qwen3.6 Plus is Alibaba's December 2025 proprietary model scoring 78.8% SWE-Bench Verified, 95.3% AIME, 90.4% GPQA Diamond, 88.5% MMLU-Pro, and 74.2% IFBench. The Plus-tier model in the Qwen3.6 API offering.
At a Glance
Where Qwen3.6 Plus leads
- 78.8% SWE-Bench Verified — competitive software engineering; near Gemini 3 Flash (78%)
- 95.3% AIME 2026 — near-perfect competition mathematics
- 90.4% GPQA Diamond — excellent science reasoning
- 88.5% MMLU-Pro — strong academic knowledge
- 74.2% IFBench — good instruction following
Where it lags
- Proprietary: no open weights
- December 2025 release — pre-dates Qwen3.5/Qwen4 generation models
- No BrowseComp/HLE data
Best for: Alibaba Cloud (DashScope) deployments requiring balanced STEM and coding; applications where Qwen API access is primary.
What Qwen3.6 Plus Is
Qwen3.6 Plus is the Plus-tier API variant in the Qwen3.6 model family, released December 2025. Unlike the open-weight Qwen3 generation (Qwen3 235B A22B, Qwen3 32B, etc.), the "Plus" API models are proprietary and available only through Alibaba's DashScope API.
The Qwen3.6 family sits between the Qwen3 (April/May 2025) and Qwen3.5 (September 2025) generations, with December 2025 being in the Qwen3.6 era of API improvements.
Specifications
| Field | Value |
|---|---|
| Organization | Alibaba |
| License | Proprietary (API only) |
| Release date | December 2025 |
| Modality | Text only |
Pricing
Available via Alibaba Cloud DashScope API. Refer to Alibaba AI pricing for current Qwen3.6 Plus rates.
Public Benchmark Scores
| Benchmark | Score | Source | Date |
|---|---|---|---|
| SWE-Bench Verified | 78.8% | Benchgen evaluation | 2025-12 |
| AIME 2026 | 95.3% | Benchgen evaluation | 2025-12 |
| GPQA Diamond | 90.4% | Benchgen evaluation | 2025-12 |
| MMLU-Pro | 88.5% | Benchgen evaluation | 2025-12 |
| IFBench | 74.2% | Benchgen evaluation | 2025-12 |
Qwen3.6 Plus vs Alternatives
| Model | SWE-Bench | GPQA Diamond | MMLU-Pro | License |
|---|---|---|---|---|
| Qwen3.6 Plus | 78.8% | 90.4% | 88.5% | Proprietary |
| Qwen3.5 122B A10B | 72% | 86.6% | 86.7% | Apache 2.0 |
| Gemini 3 Flash | 78% | 90.4% | — | Proprietary |
| Kimi K2.6 | 80.2% | 90.5% | — | Apache 2.0 |
Qwen3.6 Plus vs Qwen3.5 122B A10B: higher SWE-Bench (78.8% vs 72%) but proprietary vs open-weight. For open access: Qwen3.5 122B. For Alibaba API: Qwen3.6 Plus.
Related Models & Benchmarks
- Compare: Qwen3.5 122B A10B · Gemini 3 Flash · Kimi K2.6
- Benchmarks: SWE-Bench Verified · GPQA Diamond · MMLU-Pro
Frequently Asked Questions
What is Qwen3.6 Plus?
Alibaba's December 2025 proprietary API model scoring 78.8% SWE-Bench, 95.3% AIME, 90.4% GPQA Diamond, 88.5% MMLU-Pro. Available via DashScope.Specs from Alibaba's Qwen3.6 Plus release (December 2025) and Benchgen evaluations. Last updated 2026-07-24.