Can I run ukisai_Swift-Qwen3.8-27b-GGUF on an RTX 3060 12 GB?
Yes, just
Its IQ2_S file (9.68 GB) fits in the 10.8 GB an RTX 3060 12 GB can use, with room for ≈ 14K tokens of context. That file is under 3 bits per weight, so expect noticeably lower quality.
Full analysis of ukisai_Swift-Qwen3.8-27b-GGUF
Everything that runs on RTX 3060 12 GB
Get the IQ2_S file on Hugging Face
ollama run hf.co/bartowski/ukisai_Swift-Qwen3.8-27b-GGUF:IQ2_SFiles and context on RTX 3060 12 GB
Each file of ukisai_Swift-Qwen3.8-27b-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 3060 12 GB | Context (16-bit) | Context (8-bit) |
|---|---|---|---|---|
| BF16 | 54.7 GB | Doesn't fit | — | — |
| Q8_0 | 29.1 GB | Doesn't fit | — | — |
| Q6_K_L | 25.0 GB | Doesn't fit | — | — |
| Q6_K | 23.9 GB | Doesn't fit | — | — |
| Q6_K_S | 22.9 GB | Doesn't fit | — | — |
| Q5_K_M | 20.9 GB | Doesn't fit | — | — |
| Q5_K_S | 19.6 GB | Doesn't fit | — | — |
| Q4_K_L | 18.8 GB | Doesn't fit | — | — |
| Q4_1 | 17.8 GB | Doesn't fit | — | — |
| IQ4_NL | 17.4 GB | Doesn't fit | — | — |
| Q4_K_M | 17.4 GB | Doesn't fit | — | — |
| Q4_K_S | 16.4 GB | Doesn't fit | — | — |
| Q4_0 | 16.3 GB | Doesn't fit | — | — |
| IQ4_XS | 15.5 GB | Doesn't fit | — | — |
| IQ3_M | 14.9 GB | Doesn't fit | — | — |
| Q3_K_L | 14.1 GB | Doesn't fit | — | — |
| Q3_K_M | 13.4 GB | Doesn't fit | — | — |
| IQ3_XS | 12.8 GB | Doesn't fit | — | — |
| Q3_K_S | 12.7 GB | Doesn't fit | — | — |
| IQ3_XXS | 12.3 GB | Doesn't fit | — | — |
| Q2_K | 10.8 GB | Doesn't fit | — | — |
| IQ2_M | 10.5 GB | Doesn't fit | 2K | 4K |
| IQ2_S | 9.68 GB | Tight | 14K | 29K |
| IQ2_XS | 9.09 GB | Tight | 23K | 46K |
| IQ2_XXS | 8.88 GB | Tight | 26K | 53K |