Models / Qwen3.6 35B A3B / Calculator

Qwen3.6 35B A3B cost & VRAM calculator

What does Qwen3.6 35B A3B cost to run for your workload, and can you run it on your own hardware? Set your workload below - we compute per-provider API cost live and tell you honestly whether local hardware can run it.

Your workload

Casual Power user Team of 5 Codebase task

Qwen3.6 35B A3B runs an always-on thinking mode. Reasoning (thinking) tokens are billed at the output rate ($15.00/M), so count them here to see the thinking portion of your bill.

API cost for your workload

Provider Rate ($/1M) Monthly cost
Darkbloom $0.05 in · $0.70 out · $0.02 cache $0.0 cheapest
Venice $0.10 in · $1.00 out · $0.05 cache $0.0 cheapest
DekaLLM $0.10 in · $1.00 out · $0.05 cache $0.0 cheapest
AkashML $0.10 in · $0.90 out · $0.05 cache $0.0 cheapest
Parasail $0.10 in · $0.95 out · $0.05 cache $0.0 cheapest
Phala $0.20 in · $1.27 out · $0.06 cache $0.0 cheapest
SiliconFlow $0.24 in · $1.80 out · $0.15 cache $0.01
CoreWeave $0.25 in · $1.25 out · $0.25 cache $0.01
AtlasCloud $0.25 in · $1.25 out · $0.25 cache $0.01
DeepInfra avoid $0.10 in · $0.95 out · $0.10 cache $0.0 cheapest

Monthly cost is an estimate from list prices and your workload - verify against the provider before committing. Cached fraction applies the cache rate to that share of input.

Can you run it locally?

YES - IT RUNS ON HARDWARE YOU CAN BUY

Qwen3.6 35B A3B fits on individual rigs. Here's the hardware that runs it comfortably; the full per-quant VRAM matrix and tok/s estimates live on the model page.

4x H100 80GB (320GB) 320GB Q4_K_M
4293.7t/s
NVIDIA DGX Station 748GB 748GB Q4_K_M
2563.4t/s
8x RTX 3090 rack (192GB) 192GB Q4_K_M
2399.8t/s
4x RTX 5090 (128GB) 128GB Q4_K_M
2296.8t/s
AMD Instinct MI300X (192GB) 192GB Q4_K_M
1706.2t/s
4x RTX 4090 (96GB) 96GB Q4_K_M
1291.9t/s
2x RTX 5090 (64GB) 64GB Q4_K_M
1148.4t/s
2x RTX 3090 (48GB) 48GB Q4_K_M
600.0t/s
Single RTX 5090 (32GB) 32GB Q4_K_M
574.2t/s
RTX PRO 6000 Blackwell (96GB) 96GB Q4_K_M
574.2t/s
Mac Studio M4 Ultra 192GB 192GB Q4_K_M
381.7t/s
Mac Studio M4 Ultra 512GB 512GB Q4_K_M
381.7t/s
Single RTX 4090 (24GB) 24GB Q4_K_M
323.0t/s
MacBook Pro M5 Max 128GB 128GB Q4_K_M
214.6t/s
Dual EPYC 9004 + 768GB DDR5-4800 768GB Q4_K_M
147.7t/s
DGX Spark 128GB unified 128GB Q4_K_M
87.5t/s
Ryzen AI Max+ 395 128GB 128GB Q4_K_M
82.0t/s
Jetson AGX Orin 64GB 64GB Q4_K_M
65.6t/s
Epyc + 512GB DDR4-3200 + 2x RTX 3090 560GB Q4_K_M
65.6t/s
Epyc + 512GB DDR4-2400 + 2x RTX 3090 560GB Q4_K_M
49.2t/s
Q4_K_M
20.0GB weights 20.0GB min 22.0GB rec
Q8_0
37.0GB weights 37.0GB min 40.0GB rec

Want the own-vs-rent payback (capex vs your monthly API bill)? The budget tool computes that against representative rigs.

Full model card API pricing table Generic token calculator