SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 — Hardware Requirements & GPU Compatibility
Chat
SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 is a 785.0B-parameter open language model from OxTank in the GLM 5 family. At Q4_K_M it needs about 518.08 GB of VRAM — see which GPUs and Macs can run it below.
SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 runs on a multi-GPU server
Q4_K_M needs ~527.4 GB · about 8× H100 SXM 80 GB · or Mac Studio (M3 Ultra, 512GB) at Q3_K_M (~1 tok/s)
Macs from 512 GB (Q3_K_M · ~1 tok/s)
MIT
Not set up for it?
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
vllm serve OxTank/SuperGLM-5.2-W4W8-Abliterated-DCP4-600K-K6 --max-model-len 16384 --tensor-parallel-size 8Size the nodes from the table above
Download: Weights on Hugging Face
Other ways
each opens step-by-stepRun in cloud
Rent 8× H100 SXM node · Q4_K_M
- Estimated hourly price
- ~$21.52/hr
- est., whole server
- GPUs
- 8 GPUs
- 640 GB of VRAM
How Much VRAM Does SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | Runs on | |
|---|---|---|---|---|
| 367.0 GB | 386.8 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Runs on 6× H100 SXM 80 GB | |
| 420.9 GB | 440.8 GB | 3-bit medium — 3-bit medium quantization | Runs on 6× H100 SXM 80 GB | |
| 518.1 GB | 537.9 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Runs on 8× H100 SXM 80 GB | |
| 615.2 GB | 635.1 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Runs on 5× H200 SXM 141 GB | |
| 712.4 GB | 732.2 GB | 6-bit, very good — 6-bit quantization, very good quality | Runs on 6× H200 SXM 141 GB | |
| 863.5 GB | 883.3 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Runs on 7× H200 SXM 141 GB | |
| 1727.0 GB | 1746.8 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Runs on 4 nodes of 8× H100 SXM |
Selected Q4_K_M
4-bit medium quantization527.4GBat 16K context
- Weights
- 516.5 GB
- KV cache
- 10.6 GB
- Overhead
- 0.3 GB
Server-class: ~527.4 GB at 16K context · about 8× H100 SXM 80 GB
Which Hardware Can Run SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6?
Q4_K_M · 518.1 GBSuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 (Q4_K_M) requires 518.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 674+ GB is recommended. No consumer GPU has enough memory.
Needs server-class hardware
SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 runs on a multi-GPU server · Q4_K_M needs ~527.4 GB
| Hardware | Quant | Memory | Speed | Context | Hardware page | |
|---|---|---|---|---|---|---|
| Mac 512 GB · Mac Studio (M3 Ultra, 512GB) | Q3_K_M | 430.2 GB | ~1 tok/s | 32K ctx | Mac Studio → | Q3_K_M · 430.2 GB · ~1 tok/s · 32K ctx |
| 8× H100 SXM 80 GB · 640 GB | Q4_K_M | 527.4 GB | — | 32K ctx | H100 SXM → | Q4_K_M · 527.4 GB · 32K ctx |
| 5× H200 SXM 141 GB · 705 GB | Q4_K_M | 527.4 GB | — | 32K ctx | H200 SXM → | Q4_K_M · 527.4 GB · 32K ctx |
| 4× B200 192 GB · 768 GB | Q4_K_M | 527.4 GB | — | 32K ctx | B200 → | Q4_K_M · 527.4 GB · 32K ctx |
Also: NVIDIA DGX A100 640GB · NVIDIA DGX H100
By VRAM tier · typical GPUs and Macs at Q4_K_M
Consumer GPUs (up to 32 GB) and Macs up to 256 GB: too small for this model (needs ≥ 376.3 GB)8 GB tier →12 GB tier →16 GB tier →24 GB tier →Macs →
Alternatives to SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6
GLM 5 family · 6 models →Similar models
4| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| DeepSeek V4.1 FlashDeepSeek | 129.5 | 458.4 GB | —needs > 24 GB | same size |
| Muse Spark Open 816B A42BMeta | 125.3 | 551.9 GB | —needs > 24 GB | same size |
| Atria Dawn PreviewInternLM | — | 1508.4 GB | —needs > 24 GB | same size |
| Hy4 PreviewTencent | — | 1564.2 GB | —needs > 24 GB | same size |
Specifications
- Publisher
- OxTank
- Family
- GLM 5
- Parameters
- 785.0B
- Release Date
- 2026-07-15
- License
- MIT
Frequently Asked Questions
How much VRAM does SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 need?
SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 requires 518.1 GB of VRAM at Q4_K_M, or 1727.0 GB at BF16.
VRAM = Weights + KV Cache + Overhead
Weights = 785.0B × 4.8 bits ÷ 8 = 471 GB
KV Cache + Overhead ≈ 47.1 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
VRAM usage by quantization
Can NVIDIA GeForce RTX 5090 run SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6?
No — SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 requires at least 367.0 GB at Q2_K, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
What's the best quantization for SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6?
For SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6, Q4_K_M (518.1 GB) offers the best balance of quality and VRAM usage. Q5_K_M (615.2 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 367.0 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 on a Mac?
Yes, but only at lower quantizations. The smallest Mac that can run SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 is Mac Studio (M3 Ultra, 512GB) at Q3_K_M; 1 of the 146 Macs we list can run it at some quantization. For Q4_K_M (518.1 GB) you need a Mac with more unified memory.
Can I run SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 locally?
Yes — SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
What's the download size of SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6?
At Q4_K_M, the download is about 470.99 GB. The full-precision BF16 version is 1569.95 GB. The smallest option (Q2_K) is 333.62 GB.
Which GPUs can run SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6?
No single consumer GPU has enough VRAM to run SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 at Q4_K_M (518.1 GB). Multi-GPU or professional hardware is required.
Which devices can run SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6?
SuperGLM 5.2 W4W8 Abliterated DCP4 600K K6 requires at least 518.1 GB at Q4_K_M, which exceeds the usable memory of every unified-memory device we list (the largest has 508.5 GB).