Yes — with 45.8 GB to spare
LFM2.5 2.6B at Q4_K_M fits your Ryzen AI Max+ 395 · 64 GB entirely in unified memory at 8K context, at an estimated 94 tokens per second. There is room for its full 128K window.
Fully in unified memory
8K context
Q4_K_M · 1.5 GB
LFM Open License v1.0
Released 28 Jul 2026
New this month
Not in the Ollama library
Convolution-heavy hybrid for CPUs and NPUs: 22 of 30 blocks keep no KV cache at all.
The VRAM budget
weights 1.5 GB
Weights 1.5 GB
KV cache @ 8K 0.13 GB
Runtime overhead 0.6 GB
Free 45.8 GB of 48.0 GB
Quantisation ladder
| Quant | Weights | Total @ 8K | Max context | Tok/s | Quality | Fit |
|---|---|---|---|---|---|---|
| F16 | 5.0 GB | 5.8 GB | 128K | 28 | Reference | Long context |
| Q8_0 | 2.7 GB | 3.4 GB | 128K | 54 | −0.1% ppl | Long context |
| Q6_K | 2.1 GB | 2.8 GB | 128K | 69 | −0.4% ppl | Long context |
| Q5_K_M | 1.8 GB | 2.5 GB | 128K | 80 | −0.8% ppl | Long context |
| Q4_K_M | 1.5 GB | 2.2 GB | 128K | 94 | −1.9% ppl | Recommended |
| Q3_K_M | 1.2 GB | 2.0 GB | 128K | 116 | −5.4% ppl | Long context |
| Q2_K | 1.1 GB | 1.8 GB | 128K | 136 | −15% ppl | Long context |
Quality is the published perplexity delta against f16 weights. Max context assumes an f16 KV cache; q8_0 roughly doubles it. Only 8 of its 30 blocks keep a per-token KV cache; the rest are linear-attention, Mamba or convolution blocks with a fixed-size state.
How to run it
$ llama-server \
-hf LiquidAI/LFM2.5-2.6B:Q4_K_M \
-c 8192 -ngl 99
The engine underneath most of the others. Every knob is exposed. More on llama.cpp.
01Download is 1.5 GB. Keep it on an SSD — a first load off a spinning disk takes minutes.
02macOS caps what the GPU may wire down at about 48.0 GB of 64 GB. Raising it with iogpu.wired_limit_mb is possible, and risky.
03There is room to go to the model's full 128K context on this card.