Qwen3.5-35B-A3B
Qwen3.5-35B-A3B is an open-weights language model from Qwen, released February 24, 2026. It has 36B parameters (mixture-of-experts, 3B active). The context window is 256K tokens. On the Epoch Capabilities Index it scores 144.0, rank 102 of 256 rated models. API price: $0.25 input / $2 output per million tokens.
Check it against your own PC →
Can you run Qwen3.5-35B-A3B locally?
The smallest common GPU it runs on is 6 GB of VRAM (with 32 GB of RAM): Q6_K, GPU + experts in RAM, roughly 24 tokens per second. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Runs · GPU + CPU experts | Q6_K (≈ lossless) | 5.2 GB | 25.2 GB | ~24 |
| 8 GB VRAM | Runs · GPU + CPU experts | Q6_K (≈ lossless) | 7.2 GB | 23.2 GB | ~28 |
| 12 GB VRAM | Runs · GPU + CPU experts | Q6_K (≈ lossless) | 11.2 GB | 19.2 GB | ~34 |
| 16 GB VRAM | Runs · GPU + CPU experts | Q6_K (≈ lossless) | 15.2 GB | 15.2 GB | ~42 |
| 24 GB VRAM | Runs on GPU | Q4_K_M (very good) | 22.9 GB | – | ~230 |
| 32 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 30.4 GB | – | ~300 |
Ready-made quantizations on Hugging Face: mradermacher/Qwen3.5-35B-A3B-Base-GGUF · mradermacher/Qwen3.5-35B-A3B-Base-i1-GGUF · Mungert/Qwen3.5-35B-A3B-GGUF · maxwellb-hf/Qwen3.5-35B-A3B-Base-GGUF · Azure99/Blossom-V6.4-35B-A3B-GGUF · rachpradhan/Qwen3.5-35B-A3B-Turbo-SWE-v0.0.1
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| GPQA Diamond | 81.2% |
| OTIS Mock AIME | 70.0% |
Links
Hugging Face: Qwen/Qwen3.5-35B-A3B-Base · Epoch AI benchmarks
Models with a similar score
Qwen3.5 397B-A17B · Qwen3.7-Plus · MiniMax-M3 · Claude Sonnet 4.5 · o3-pro · Qwen 3.5 Plus (hosted 397B-A17B) · o3 · Nemotron 3 Ultra
More from Qwen
Qwen3.8 Omni Flash · Qwen3.8 Max (0902) · Qwen3.8 Flash · Qwen3.8 Flash Next · Qwen 3.8 27B · Qwen3.8-2.4T-A95B · Qwen 3.8 Max · Qwen3.7 Flash