No — not on this device
Kimi K2.6 1T-A32B at Q4_K_M needs 578.6 GB against 204.8 GB usable, and the shortfall of 373.8 GB is more than 32 GB of system RAM can cover at a tolerable speed. A smaller sibling or a lower quantisation is the honest answer here.
Does not fit
8K context
Q4_K_M · 577.5 GB
Modified MIT
Released 20 Apr 2026
Not in the Ollama library
The open coding-agent benchmark leader of spring 2026 (80.2 SWE-bench). A 512 GB Mac Studio pair, or a ceiling.
The VRAM budget
weights 577.5 GB
Weights 577.5 GB
KV cache @ 8K 0.54 GB
Runtime overhead 0.6 GB
Over budget 373.8 GB past 204.8 GB
Quantisation ladder
| Quant | Weights | Total @ 8K | Max context | Tok/s | Quality | Fit |
|---|---|---|---|---|---|---|
| Q8_0 | 1016.2 GB | 1017.4 GB | — | ~1.8 | −0.1% ppl | 812.6 GB over |
| Q6_K | 784.3 GB | 785.4 GB | — | ~2.3 | −0.4% ppl | 580.6 GB over |
| Q5_K_M | 677.9 GB | 679.0 GB | — | ~2.7 | −0.8% ppl | 474.2 GB over |
| Q4_K_M | 577.5 GB | 578.6 GB | — | ~3.2 | −1.9% ppl | 373.8 GB over |
| Q3_K_M | 467.5 GB | 468.6 GB | — | ~3.9 | −5.4% ppl | 263.8 GB over |
| Q2_K | 400.5 GB | 401.7 GB | — | ~4.6 | −15% ppl | 196.9 GB over |
Quality is the published perplexity delta against f16 weights. Max context assumes an f16 KV cache; q8_0 roughly doubles it. This model uses multi-head latent attention, so its cache is a compressed latent rather than full K and V.
How to run it
$ llama-server \
-hf moonshotai/Kimi-K2.6:Q4_K_M \
-c 8192 -ngl 21
The engine underneath most of the others. Every knob is exposed. More on llama.cpp.
01Download is 577.5 GB. Keep it on an SSD — a first load off a spinning disk takes minutes.
02With no GPU, thread count matters more than clock. Start at one thread per physical core.
03If it falls back to CPU silently, drop the context first, then step down a quantisation.