How much VRAM does SeedVR2_comfyUI need?

Q4_K_M is exactly 11,511,957,408 bytes (10.72 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

numz · text

SeedVR2_comfyUI

numz/SeedVR2_comfyUI

SeedVR2_comfyUI at Q4_K_M is exactly 11,511,957,408 bytes (10.72 GiB / 11.51 GB) — an effective 27.155 bits per weight, not the nominal 4.

From the file· summed from 3 file(s)

Parameters

3.4B

Architecture

seedvr

Context

—

native (config.json)

License

apache-2.0

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q3_K_M3 shards	8.29 GiB	8,903,786,400	21.003	—	cmeka
Q4_K_M3 shards	10.72 GiB	11,511,957,408	27.155	—	cmeka
Q5_K_M3 shards	13.01 GiB	13,966,706,592	32.945	—	cmeka
Q6_K3 shards	15.44 GiB	16,574,877,600	—	—	cmeka
Q8_03 shards	19.87 GiB	21,330,954,144	—	—	cmeka

Compare with

same modality, comparable size

vs Qwen3-Coder-30B-A3B-Instruct vs Qwen3.6-27B vs Qwen3.8-27B vs gemma-4-12B-it-qat-q4_0-unquantized vs gemma-4-E4B-it vs Qwen3-30B-A3B-Thinking-2507 All alternatives →Will it fit on my card? →

Why other calculators give a different number

A parameters × bits ÷ 8 estimate puts Q4_K_M at roughly 1.78 GiB. The real file is 10.72 GiB, because a quantization is a mixture and some tensors are always kept at higher precision.

Architecture

Architecture unavailable — this repository is gated and no ungated mirror was found. Exact file sizes above are still authoritative; only the KV math needs the config.

Questions people ask

How much VRAM does SeedVR2_comfyUI need?: Q4_K_M is exactly 11,511,957,408 bytes (10.72 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of SeedVR2_comfyUI 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.