VAETKI
VAETKI is an open-weights language model from NC AI, released December 30, 2025. It has 112B parameters (mixture-of-experts, 10B active). The context window is 128K tokens.
Check it against your own PC →
Lab
NC AI
Released
December 30, 2025
Weights
Open mit
Parameters
112B MoE · 10B active
Context
128K
Price / M tokens
–
Epoch ECI
not yet rated
Can you run VAETKI locally?
The smallest common GPU it runs on is 6 GB of VRAM (with 32 GB of RAM): IQ2_XXS, GPU + experts in RAM, roughly 16 tokens per second. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Runs · GPU + CPU experts | IQ2_XXS (heavy loss) | 5.2 GB | 24.5 GB | ~16 |
| 8 GB VRAM | Runs · GPU + CPU experts | IQ2_XXS (heavy loss) | 7.2 GB | 22.5 GB | ~17 |
| 12 GB VRAM | Runs · GPU + CPU experts | IQ2_XXS (heavy loss) | 11.2 GB | 18.5 GB | ~21 |
| 16 GB VRAM | Runs · GPU + CPU experts | IQ2_M (heavy loss) | 15.2 GB | 23.4 GB | ~17 |
| 24 GB VRAM | Runs · GPU + CPU experts | IQ3_XXS (noticeable loss) | 23.2 GB | 20.5 GB | ~20 |
| 32 GB VRAM | Runs · GPU + CPU experts | Q3_K_M (okay) | 31.2 GB | 24.4 GB | ~18 |
Benchmark results
Epoch AI has not published results for VAETKI yet. New models are usually evaluated within days to weeks.