Can I run it?
Can I run Qwen3.8 Flash Next on GeForce RTX 3090?
Won't fit
Won't fit on GeForce RTX 3090. The smallest quant UD-IQ1_S is 67.6 GB before KV cache, and usable memory is 22.3 GB.
Best quantUD-IQ1_S
Working context—
Max that fits—
Speed—
Weights67.6 GB
KV cache0.00 GB
Total0.00 GB
Usable22.3 GB
Qwen3.8 Flash Next details · GeForce RTX 3090 page · How this estimate is made
Run commands
llama.cpp
llama-server -hf unsloth/Qwen3.8-Flash-Next-GGUF:UD-IQ1_S -c 2048 -ngl 0
Ollama
ollama run qwen3.8-flash-next
LM Studio
lms get unsloth/Qwen3.8-Flash-Next-GGUF # Then load quant UD-IQ1_S
Assumptions
- Estimate, not a benchmark. Decode speed uses a memory-bandwidth model at 55–75% of peak.
- 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%.
- 0.5 GB is reserved because the GPU also drives a display, then 5% is left free.
- Usable memory for this profile is 22.3 GB.
Models that fit this machine more comfortably
- Can I run Qwen3 Next 80B A3B Thinking on GeForce RTX 3090? — Tight fit, UD-IQ1_S, 487–664 tok/s
- Can I run DeepSeek R1 Distill Llama 70B on GeForce RTX 3090? — Tight fit, UD-IQ2_XXS, 23–32 tok/s
- Can I run Llama 3.3 70B Instruct on GeForce RTX 3090? — Tight fit, UD-IQ2_XXS, 23–32 tok/s
Updated 2026-10-24. Profile on this page is the device default plus 32 GB of DDR5-5600 system RAM.