Can I run gemma-4-26B-A4B-it-GGUF on an RTX 3060 12 GB?
Yes, just
Its UD-IQ2_M file (10.0 GB) fits in the 10.8 GB an RTX 3060 12 GB can use, with room for ≈ 4K tokens of context. That file is under 3 bits per weight, so expect noticeably lower quality.
Full analysis of gemma-4-26B-A4B-it-GGUF
Everything that runs on RTX 3060 12 GB
Get the UD-IQ2_M file on Hugging Face
ollama run hf.co/unsloth/gemma-4-26B-A4B-it-GGUF:UD-IQ2_MFiles and context on RTX 3060 12 GB
Each file of gemma-4-26B-A4B-it-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 | 50.5 GB | Doesn't fit | — | — |
| UD-Q8_K_XL | 27.6 GB | Doesn't fit | — | — |
| Q8_0 | 26.9 GB | Doesn't fit | — | — |
| UD-Q6_K_XL | 23.3 GB | Doesn't fit | — | — |
| UD-Q6_K | 23.2 GB | Doesn't fit | — | — |
| UD-Q5_K_XL | 21.2 GB | Doesn't fit | — | — |
| UD-Q5_K_M | 21.2 GB | Doesn't fit | — | — |
| UD-Q5_K_S | 18.9 GB | Doesn't fit | — | — |
| UD-Q4_K_XL | 17.0 GB | Doesn't fit | — | — |
| UD-Q4_K_M | 16.9 GB | Doesn't fit | — | — |
| MXFP4_MOE | 16.6 GB | Doesn't fit | — | — |
| UD-Q4_K_S | 16.5 GB | Doesn't fit | — | — |
| UD-IQ4_NL | 13.6 GB | Doesn't fit | — | — |
| UD-IQ4_XS | 13.6 GB | Doesn't fit | — | — |
| UD-Q3_K_XL | 12.9 GB | Doesn't fit | — | — |
| UD-Q3_K_M | 12.7 GB | Doesn't fit | — | — |
| UD-IQ3_XXS | 11.4 GB | Doesn't fit | — | — |
| UD-IQ3_S | 11.3 GB | Doesn't fit | — | — |
| UD-Q2_K_XL | 10.5 GB | Doesn't fit | 514 | 1K |
| UD-IQ2_M | 10.0 GB | Tight | 4K | 14K |
| UD-IQ2_XXS | 9.92 GB | Tight | 5K | 16K |