More
Submit build

Models / Qwen3.5 35B-A3B

Best home rig for Qwen3.5 35B-A3B

Quick answer

How much VRAM does Qwen3.5 35B-A3B need?
The weights are about 22.0 GB at Q4_K_M (35.95B params × 4.89 bits ÷ 8, before context/KV cache). Add memory for context (KV cache) and the runtime.

What is the fastest measured hardware for Qwen3.5 35B-A3B?
ThinkPad X1 Carbon Gen 13 (Core Ultra 7 258V, Arc 140V iGPU, 32 GB): Qwen3.5 35B-A3B IQ4_XS at 25.6 tok/s on llama.cpp Vulkan, 22.3 at 8K depth: 25.58 tok/s generation (llama.cpp 12377 Vulkan (Mesa ANV), IQ4_XS (16.95 GiB)), from https://github.com/ggml-org/llama.cpp/discussions/10879.

Last updated:

Legacy model35.95B total · 3B active4 measured resultsfastest plain decode: 25.58 tok/s model info ↗

Newer model available: Qwen3.6 35B-A3B. Qwen3.5 35B-A3B is a legacy model; its numbers stay here for reference. same-size Qwen MoE, newer release. source ↗

Real, measured generation speeds for Qwen3.5 35B-A3B on home hardware, fastest first. Different quants/engines aren't directly comparable.

BuildQuantEnginetok/sPrompt processing (tok/s)CostSource
ThinkPad X1 Carbon Gen 13 (Core Ultra 7 258V, Arc 140V iGPU, 32 GB): Qwen3.5 35B-A3B IQ4_XS at 25.6 tok/s on llama.cpp Vulkan, 22.3 at 8K depthIQ4_XS (16.95 GiB)llama.cpp 12377 Vulkan (Mesa ANV)25.58379 @8192 · est. TTFT @8k ≈21.6 s*github.com ↗
Radeon 890M iGPU (Ryzen AI 9 HX 370): Qwen3.5 35B-A3B Q4_K_M at 24.8 tok/s on llama.cpp Vulkan b10830, 20.3 on b12387 (tg regression report)Unsloth UD-Q4_K_Mllama.cpp b10830 465e49b9c (Vulkan, RADV)24.8251 @512 · est. TTFT @8k ≈32.6 s*github.com ↗
NVIDIA Jetson AGX Xavier 32 GB (Volta SM72): LFM2.5 24B-A2B Q4_K_M at 36.9 tok/s and Qwen3.5 35B-A3B Q4_K_M at 24.6 on llama.cpp v1.0.0 CUDAQ4_K_M (20.21 GiB)llama.cpp v1.0.0 11fe021 (CUDA 11.4, SM72)24.6228 @512 · est. TTFT @8k ≈35.9 s*github.com ↗
Radeon 890M iGPU (Ryzen AI 9 HX 370): Qwen3.5 35B-A3B Q4_K_M at 24.8 tok/s on llama.cpp Vulkan b10830, 20.3 on b12387 (tg regression report)Unsloth UD-Q4_K_Mllama.cpp b12387 a956a852c (Vulkan, RADV)20.32293 @512 · est. TTFT @8k ≈28.0 s*github.com ↗

Prompt processing = prefill speed reported by the source, @ prompt length. * est. TTFT @8k is calculated from measured prefill, not measured. Why it matters.

Rent before you buy: try this model on a cloud GPU for a few dollars first: Runpod (referral link; sign up with Google to get the new-user bonus)

Memory math: how much VRAM / RAM does Qwen3.5 35B-A3B need?

Qwen3.5 35B-A3B has 35.95B parameters, about 3B active per token (source: Qwen model card: 35B total / 3B activated MoE (8 routed + 1 shared experts), Apache 2.0, HF repo created 24 Feb 2026; HF safetensors total 35.95B (includes vision encoder)). Weight size = parameters × bits-per-weight ÷ 8, before context (KV cache) and runtime overhead.

FormatBits/weightWeightsSmallest common memory size that holds the weights alone
Q8_08.5038.2 GB48 GB
Q6_K6.5629.5 GB32 GB
Q5_K_M5.7025.6 GB32 GB
Q4_K_M4.8922.0 GB24 GB
Q3_K_M4.0018.0 GB24 GB
Q2_K3.1614.2 GB16 GB

Bits-per-weight for GGUF quants are llama.cpp's own figures (llama.cpp quantize README). "As measured" rows apply the source's stated bits-per-weight to the parameter count; real files differ slightly because some layers stay at higher precision. Try other combinations in the VRAM calculator.

Why it can still run on smaller GPUs: it is a mixture-of-experts model, so only about 3B of its 35.95B parameters are used for each generated token. Runtimes such as llama.cpp (--n-cpu-moe), ExLlamaV3 or Project Maya can keep most expert weights in system RAM and only the shared layers and hot experts in VRAM. Then you need VRAM + system RAM ≥ weights, and speed is limited mainly by RAM bandwidth.

FAQ

How much VRAM does Qwen3.5 35B-A3B need?

The weights are about 22.0 GB at Q4_K_M (35.95B params × 4.89 bits ÷ 8, before context/KV cache). Add memory for context (KV cache) and the runtime.

What is the fastest measured hardware for Qwen3.5 35B-A3B?

ThinkPad X1 Carbon Gen 13 (Core Ultra 7 258V, Arc 140V iGPU, 32 GB): Qwen3.5 35B-A3B IQ4_XS at 25.6 tok/s on llama.cpp Vulkan, 22.3 at 8K depth: 25.58 tok/s generation (llama.cpp 12377 Vulkan (Mesa ANV), IQ4_XS (16.95 GiB)), from https://github.com/ggml-org/llama.cpp/discussions/10879.

Related: what GPU do you need for 2026 models · all builds