Can I run Qwen3.5-35B-A3B-GGUF on an RTX 4090?
Yes, just
Its Q4_K_S file (20.7 GB) fits in the 21.6 GB an RTX 4090 can use, with room for ≈ 41K tokens of context.
Full analysis of Qwen3.5-35B-A3B-GGUF
Everything that runs on RTX 4090
Get the Q4_K_S file on Hugging Face
ollama run hf.co/unsloth/Qwen3.5-35B-A3B-GGUF:Q4_K_SFiles and context on RTX 4090
Each file of Qwen3.5-35B-A3B-GGUF (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.
| File | Size | On RTX 4090 | Context (16-bit) | Context (8-bit) |
|---|---|---|---|---|
| BF16 | 69.4 GB | Doesn't fit | — | — |
| UD-Q8_K_XL | 48.7 GB | Doesn't fit | — | — |
| Q8_0 | 36.9 GB | Doesn't fit | — | — |
| UD-Q6_K_XL | 32.1 GB | Doesn't fit | — | — |
| Q6_K | 28.9 GB | Doesn't fit | — | — |
| UD-Q6_K_S | 28.5 GB | Doesn't fit | — | — |
| UD-Q5_K_XL | 26.4 GB | Doesn't fit | — | — |
| Q5_K_M | 26.2 GB | Doesn't fit | — | — |
| Q5_K_S | 24.8 GB | Doesn't fit | — | — |
| UD-Q4_K_XL | 22.2 GB | Doesn't fit | — | — |
| Q4_K_M | 22.0 GB | Doesn't fit | — | — |
| MXFP4_MOE | 21.6 GB | Doesn't fit | — | — |
| Q4_K_S | 20.7 GB | Tight | 41K | 82K |
| UD-Q4_K_L | 20.2 GB | Tight | 63K | 127K |
| UD-IQ4_NL | 17.8 GB | Tight | 177K | 256K |
| UD-IQ4_XS | 17.5 GB | Tight | 193K | 256K |
| UD-Q3_K_XL | 16.6 GB | Fits | 235K | 256K |
| Q3_K_M | 16.4 GB | Fits | 247K | 256K |
| Q3_K_S | 15.3 GB | Fits | 256K | 256K |
| UD-IQ3_S | 13.6 GB | Fits | 256K | 256K |
| UD-IQ3_XXS | 13.1 GB | Fits | 256K | 256K |
| UD-Q2_K_XL | 12.2 GB | Fits | 256K | 256K |
| UD-IQ2_M | 11.4 GB | Fits | 256K | 256K |
| UD-IQ2_XXS | 10.7 GB | Fits | 256K | 256K |