How much VRAM does SuperHY3-abliterated-NVFP4 need?

IQ2_M is exactly 102,116,475,744 bytes (95.10 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

How large is SuperHY3-abliterated-NVFP4's KV cache?

10.00 GiB at 32K context with an f16 cache, computed per layer from the model's architecture. Quantizing the cache to q8_0 roughly halves it.

Is SuperHY3-abliterated-NVFP4 a mixture-of-experts model?

Yes — 192 experts with 8 routed per token. Every expert must be resident in memory, but only the routed ones are read per token, so its memory requirement and its speed behave very differently.

Jiunsong · text · mixture of experts

SuperHY3-abliterated-NVFP4

Jiunsong/SuperHY3-abliterated-NVFP4

SuperHY3-abliterated-NVFP4 at IQ2_M is exactly 102,116,475,744 bytes (95.10 GiB / 102.12 GB) — an effective 4.751 bits per weight, not the nominal 2. Its KV cache at 32K is 10.00 GiB.

From the file· summed from 3 file(s)From the file· KV per layer

Parameters

172B

total, not active

Architecture

hy-v3

80 layers

Context

262,144

native (config.json)

License

apache-2.0

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
IQ2_M3 shards	95.10 GiB	102,116,475,744	4.751	—	Jiunsong

KV cache by context

computed per layer

Context	KV cache (f16)●	Flat formula	Overstated by	Full / windowed / recurrent
4,096	1.25 GiB	1.25 GiB	—	80 / 0 / 0
8,192	2.50 GiB	2.50 GiB	—	80 / 0 / 0
16,384	5.00 GiB	5.00 GiB	—	80 / 0 / 0
32,768	10.00 GiB	10.00 GiB	—	80 / 0 / 0
65,536	20.00 GiB	20.00 GiB	—	80 / 0 / 0
131,072	40.00 GiB	40.00 GiB	—	80 / 0 / 0

Compare with

same modality, comparable size

vs Qwen3-Coder-30B-A3B-Instruct vs DeepSeek-V4-Flash vs Hy3 vs Qwen-AgentWorld-35B-A3B vs GLM-5.2 vs Ornith-1.0-35B All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts IQ2_M at roughly 90.08 GiB. The real file is 95.10 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

from config.json

Layers

Attention heads

KV heads

Head dim

128

Hidden size

4096

Vocab

120,832

Sliding window

none

SWA period

—

MLA

Experts

192

Experts per token

use_sliding_window

—

Questions people ask

How much VRAM does SuperHY3-abliterated-NVFP4 need?: IQ2_M is exactly 102,116,475,744 bytes (95.10 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
How large is SuperHY3-abliterated-NVFP4's KV cache?: 10.00 GiB at 32K context with an f16 cache, computed per layer. Quantizing the cache to q8_0 roughly halves it, which is often the difference between a context length fitting and not.
Is SuperHY3-abliterated-NVFP4 a mixture-of-experts model?: Yes — 192 experts, 8 routed per token. Every expert must be resident, but only the routed ones are read per token, which is why its memory requirement and its speed behave very differently.
Which quantization of SuperHY3-abliterated-NVFP4 should I use?: Q4_K_M is the usual default. Pick the largest quantization that fits your card at the context you actually need — the table above gives exact sizes for every one published.