How much VRAM does GLM-5.2-ABLITERATED-BF16 need?

Q4_K_M is exactly 454,660,890,496 bytes (423.44 GiB) in weights. Add the KV cache, which depends on context length, plus roughly half a gigabyte of runtime overhead.

Blackfrost-AI · text

GLM-5.2-ABLITERATED-BF16

Blackfrost-AI/GLM-5.2-ABLITERATED-BF16

GLM-5.2-ABLITERATED-BF16 at Q4_K_M is exactly 454,660,890,496 bytes (423.44 GiB / 454.66 GB) — an effective 4.828 bits per weight, not the nominal 4.

From the file· summed from 10 file(s)

Parameters

753B

Architecture

glm-dsa

Context

—

native (config.json)

License

other

Shipped quantizations

● exact bytes, summed from published files

Quant	Size●	Exact bytes●	Effective bpw●	Tensors●	Publisher
Q4_K_M10 shards	423.44 GiB	454,660,890,496	4.828	—	frz1
Q5_K_M11 shards	497.39 GiB	534,072,139,936	5.672	—	frz1
Q6_K13 shards	576.09 GiB	618,568,907,008	6.569	—	frz1
Q8_017 shards	745.85 GiB	800,848,098,144	8.505	—	frz1

Compare with

same modality, comparable size

vs Laguna-S-2.1 vs Hy3 vs GLM-5.2 vs Qwen3-235B-A22B vs DeepSeek-V3-0324 vs MiniMax-M2.7 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 394.65 GiB. The real file is 423.44 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 GLM-5.2-ABLITERATED-BF16 need?: Q4_K_M is exactly 454,660,890,496 bytes (423.44 GiB) in weights. Add the KV cache, which depends on your context length, plus roughly half a gigabyte of runtime overhead.
Which quantization of GLM-5.2-ABLITERATED-BF16 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.