Can I run it?

Can I run Qwen2.5 32B Instruct on GeForce RTX 3080 10GB?

Partial offload

Partial offload on your GeForce RTX 3080 10GB: Q4_K_M at 4K context, 26 of 64 layers on the GPU and 10.9 GB in system RAM. Expect about 3.8–5.2 tok/s.

Qwen2.5 32B Instruct details · GeForce RTX 3080 10GB page · How this estimate is made

Run commands

llama-server -hf bartowski/Qwen2.5-32B-Instruct-GGUF:Q4_K_M -c 4096 -ngl 26

ollama run qwen2.5

lms get bartowski/Qwen2.5-32B-Instruct-GGUF
# Then load quant Q4_K_M

Assumptions

• Estimate from memory bandwidth, not a benchmark. Decode speed uses 55–75% of peak bandwidth.
• KV cache is FP16. Engine overhead assumes llama.cpp (0.3 GB + 1% of the weights). Ollama is closer to 0.9 GB + 3%, LM Studio closer to 1 GB + 3%.
• Q4_K_M size is the real GGUF file size from Hugging Face.
• 0.5 GB is reserved because this GPU also drives a display, then 5% of the rest is left free.
• Speed is quoted at 4K context, not at the maximum context that still fits.
• About 26 of 64 layers stay on the GPU. 10.9 GB is read from system RAM.

Models that fit this machine more comfortably

• Qwen3 32B — Heavily degraded, UD-IQ1_S, 47–65 tok/s (estimate)
• Qwen3 VL 32B Instruct — Heavily degraded, UD-IQ1_S, 47–65 tok/s (estimate)
• GLM 4 32B 0414 — Heavily degraded, UD-IQ1_M, 48–65 tok/s (estimate)

Updated 2026-10-24. Profile on this page is the device default plus 32 GB of DDR5-5600 system RAM.