Can I Run Nex-N2.5 on 24 GB system RAM?

作者： Jakub Rusinowski · 最后更新： 2026年9月8日

Yes, but it is tight

Yes, but it is tight — Nex-N2.5 mini at Q3_K_M needs about 17.6 GB of the 19.2 GB usable on 24 GB system RAM, leaving only ~1.6 GB before the runtime starts swapping. Expect ~28.4 tok/s (estimated), with room for about 8,192 tokens of context.

Confidence: low · Recommended quantization: Q3_K_M · Estimated speed: ~28.4 tok/s

Get a personalized upgrade path →

购买此硬件 AMD Radeon RX 7900 XT 20GB — 20 GB VRAM · 315 W board power 立即云端部署 RunPod 上的 RTX 4090 — 低至 $0.34/小时 · 价格核实于 2026-07

或在 Vast.ai 比较，低至 $0.35/小时 (typical low · varies)

作为亚马逊联盟成员，我们从符合条件的购买中获得收入。云 GPU 链接为推荐链接——我们可能获得佣金，您无需额外付费。

24 GB system RAM — what it gives a model

Usable memory for models	19.2 GB
Memory bandwidth	90 GB/s

Nex-N2.5 on 24 GB system RAM: memory by quantization

Quant	Memory needed	Fits 19.2 GB?	Max context	Est. speed	Download
F16	72.7 GB	✗ No	—	—	70 GB
Q8_0	39.9 GB	✗ No	—	—	37.2 GB
Q6_K	31.4 GB	✗ No	—	—	28.7 GB
Q5_K_M	27.5 GB	✗ No	—	—	24.8 GB
Q4_K_M	23.8 GB	✗ No	—	—	21.1 GB
Q3_K_M	17.6 GB	✓ Yes	8K	~28.4 tok/s	14.9 GB
Q2_K	14.2 GB	✓ Yes	16K	~32.4 tok/s	11.5 GB

Which Nex-N2.5 sizes fit

Variant	Needs at Q4_K_M	Fits?	Est. speed
Nex-N2.5 Max	967.4 GB	✗ Too large	—
Nex-N2.5 Pro	244.7 GB	✗ Too large	—
Nex-N2.5 mini	23.8 GB	✗ Too large	—

What to watch out for

Q3_K_M is the only quantization that fits, and it is a heavily degraded one — expect noticeably worse output than the same model at Q4_K_M. A smaller model at Q4 is usually the better trade.
3 larger variants of Nex-N2.5 do not fit and would need CPU offload or different hardware.
This model does not publish its full attention configuration, so the KV-cache share of these figures is inferred from its parameter count rather than computed exactly.
These figures assume CPU-only inference. Any discrete GPU, even an 8 GB one, will be several times faster for models that fit in its VRAM.

Recommended setup

llama.cpp (CPU build) or Ollama — both run without a GPU

How these numbers are calculated

Memory need = quantized weights + KV cache at 8,192 tokens (f16) + 0.8 GB runtime overhead.
19.2 GB of the 24 GB is treated as usable for model weights (80% — the rest is the OS and running applications).
DDR5-5600 dual channel at 89.6 GB/s peak. CPU decode is assumed to sustain 35% of that peak, because CPU inference is not purely bandwidth-bound — it also spends real time in compute and thread synchronisation. This figure is an assumption, not a fitted constant: no CPU measurement is in the calibration set.
CPU-only inference: no GPU is assumed. A GPU of any size will beat these figures substantially.
Throughput is a memory-bandwidth roofline estimate, not a measurement. It is labelled "estimated" everywhere it appears.
KV cache is inferred from the parameter count because this model does not publish its full attention configuration.

FAQ