Can I run Qwen3.8-27B-pi-GGUF on an RTX 5070?
Yes, just
Its IQ2_M file (10.0 GB) fits in the 10.8 GB an RTX 5070 can use, with room for ≈ 10K tokens of context. That file is under 3 bits per weight, so expect noticeably lower quality.
Full analysis of Qwen3.8-27B-pi-GGUF
Everything that runs on RTX 5070
Get the IQ2_M file on Hugging Face
ollama run hf.co/bytkim/Qwen3.8-27B-pi-GGUF:IQ2_MFiles and context on RTX 5070
Each file of Qwen3.8-27B-pi-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 5070 | Context (16-bit) | Context (8-bit) |
|---|---|---|---|---|
| BF16 | 53.8 GB | Doesn't fit | — | — |
| Q8_0 | 28.6 GB | Doesn't fit | — | — |
| Q6_K | 22.1 GB | Doesn't fit | — | — |
| Q5_K_M | 19.2 GB | Doesn't fit | — | — |
| Q5_K_S | 18.7 GB | Doesn't fit | — | — |
| Q4_K_M | 16.5 GB | Doesn't fit | — | — |
| IQ4_NL | 15.8 GB | Doesn't fit | — | — |
| Q4_K_S | 15.6 GB | Doesn't fit | — | — |
| IQ4_XS | 15.1 GB | Doesn't fit | — | — |
| Q3_K_L | 14.3 GB | Doesn't fit | — | — |
| Q3_K_M | 13.3 GB | Doesn't fit | — | — |
| IQ3_M | 12.6 GB | Doesn't fit | — | — |
| Q3_K_S | 12.1 GB | Doesn't fit | — | — |
| IQ3_XS | 12.0 GB | Doesn't fit | — | — |
| Q2_K | 10.7 GB | Doesn't fit | — | — |
| IQ2_M | 10.0 GB | Tight | 10K | 19K |
| IQ2_S | 9.36 GB | Tight | 19K | 38K |
| IQ2_XS | 9.09 GB | Tight | 23K | 46K |