Alibaba
Qwen 3.5 35B A3B
最先端
Jun 2025公開日131K トークンコンテキストApache 2.0ライセンス97 卓越品質
Qwen 3.5 35B A3B (35B parameters) requires approximately 24.6 GB of VRAM with Q4_K_M quantization. As a Mixture of Experts model with 3B active parameters, it uses less memory than its total parameter count suggests. For the best balance of quality and speed, we recommend hardware with at least 29 GB of VRAM.
はじめに
— コピー&ペーストでローカル実行Copy-paste commands to run Qwen 3.5 35B A3B on your machine.
Run
ollama run qwen3.5:35b-a3bQuick specs
Parameters35B (3B active)
Architecturemoe (MoE)
Context131K tokens
Modalitytext
Min RAM13.7 GB
Rec. RAM21.3 GB (Q4_K_M)
LicenseApache 2.0
FamilyQwen
✓ Code✓ Chat
About this model
- •MoE architecture: 35B total params but only 3B active — fast and memory-efficient
- •Quality comparable to dense 7-8B models at a fraction of the compute
- •Runs well on 8 GB VRAM GPUs at Q4_K_M
関連モデル
おすすめ
最適なハードウェア
Qwen 3.5 35B A3Bのおすすめ
このモデルを実行
量子化オプション
量子化レベル別VRAM推定値
No hardware detected — fit column shows raw VRAM estimates
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 13.7 GB | Low | — |
Q3_K_S | 3 | 17.2 GB | Low | — |
NVFP4 | 4 | 19.6 GB | Medium | — |
Q4_K_M | 4 | 21.3 GB | Medium | — |
Q5_K_M | 5 | 25.2 GB | High | — |
Q6_K | 6 | 28.7 GB | High | — |
Q8_0 | 8 | 37.5 GB | Very High | — |
F16 | 16 | 71.8 GB | Maximum | — |
Quality benchmarks
Qwen 3.5 35B A3B benchmark scores
Coding
SWE-bench Verified69.2%
HumanEval+—
Aider Polyglot—
LiveCodeBench74.6%
Reasoning
MMLU-Pro85.3%
GPQA Diamond84.2%
MATH-500—
ARC Challenge—
General
Chatbot Arena—
IFEval91.9%
Source: official · 2026-03-01
ハードウェア互換性
全ハードウェアの適合度推定
Computing compatibility...
メモリ内訳
Reference: RTX 2060 6GB
Weights21.3 GB
KV Cache1.5 GB
Runtime1.2 GB
Headroom0.6 GB
よくある質問
FAQ — Qwen 3.5 35B A3B
関連項目