GLM 5.3 Flash MXFP4 — Hardware Requirements & GPU Compatibility
Chat
GLM 5.3 Flash MXFP4 is a 180.0B-parameter open language model from OneNexus in the GLM 5 family. It supports a context window of up to 1,048,576 tokens. At Q4_K_M it needs about 109.81 GB of VRAM — see which GPUs and Macs can run it below.
GLM 5.3 Flash MXFP4 runs on a 128 GB+ Mac or a datacenter GPU
Best local pick → Mac Pro M2 Ultra (128 GB) · Q4_K_M · 120.4 GB · ~5 tok/s
Macs from 96 GB (Q2_K · ~7 tok/s)
MIT
Not set up for it?
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run GLM 5.3 Flash MXFP4
One recommended way. Every other method, OS and quant is in the step-by-step guide.
Recommended
vllm serve OneNexus/GLM-5.3-Flash-MXFP4 --max-model-len 16384 --tensor-parallel-size 8vLLM repo with its own quantisation
Download: Weights on Hugging Face
Other ways
each opens step-by-stepRun in cloud
Fits on B200 (192 GB) (71 GB headroom) · Q4_K_M
- Generation speed
- ~47 tok/s
- generation speed
- Cost per 1M output tokens
- $35.04
- per 1M output tokens
How Much VRAM Does GLM 5.3 Flash MXFP4 Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | Runs on | |
|---|---|---|---|---|
| 78.3 GB | 101 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Runs on RTX PRO 6000 Blackwell Max-Q Workstation Edition 96 GB | |
| 89.6 GB | 112.2 GB | 3-bit medium — 3-bit medium quantization | Runs on Instinct MI250X 128 GB | |
| 109.8 GB | 132.5 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Runs on Instinct MI250X 128 GB | |
| 130.1 GB | 152.7 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Runs on H200 NVL 141 GB | |
| 150.3 GB | 173 GB | 6-bit, very good — 6-bit quantization, very good quality | Runs on B200 192 GB | |
| 181.8 GB | 204.5 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Runs on Instinct MI325X 256 GB | |
| 361.8 GB | 384.4 GB | 16-bit · bfloat — Brain floating point 16 — preferred for training | Runs on 6× H100 SXM 80 GB |
Selected Q4_K_M
4-bit medium quantization120.4GBat 16K context
- Weights
- 108.0 GB
- KV cache
- 12.1 GB
- Overhead
- 0.3 GB
Server-class: ~120.4 GB at 16K context · one Instinct MI250X 128 GB
Which Hardware Can Run GLM 5.3 Flash MXFP4?
Q4_K_M · 109.8 GBGLM 5.3 Flash MXFP4 (Q4_K_M) requires 109.8 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 143+ GB is recommended. Using the full 1049K context window can add up to 771.6 GB, bringing total usage to 881.4 GB. No consumer GPU has enough memory. 18 devices with unified memory can run GLM 5.3 Flash MXFP4, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), Mac Pro M2 Ultra (128 GB).
96 GB
Q4_K_M · best fit Q2_K (88.9 GB)
3 GPUs · 2 Mac classes
RTX PRO 6000 Blackwell Max-Q Workstation Edition~15Simulate output on RTX PRO 6000 Blackwell Max-Q Workstation EditionRTX PRO 6000 Blackwell Workstation Edition~15Simulate output on RTX PRO 6000 Blackwell Workstation EditionRTX PRO 6000 Blackwell Server Edition~13Simulate output on RTX PRO 6000 Blackwell Server EditionM3 Ultra*96 GB~7Simulate output on Mac Studio (M3 Ultra, 96GB)M3 Max*96 GB~3Simulate output on MacBook Pro 14" M3 Max (96 GB)128–144 GB
fits · 7.6 GB free · up to 16K ctx
4 GPUs · 2 Mac classes · 2 device families
GH200 Grace Hopper Superchip~29Simulate output on GH200 Grace Hopper SuperchipH200 NVL~28Simulate output on H200 NVLH200 SXM~28Simulate output on H200 SXMM2 Ultra*128 GB~5Simulate output on Mac Pro M2 Ultra (128 GB)M5 Max128 GB~4Simulate output on Mac Studio M5 Max (128 GB)GB10~2Simulate output on ASUS Ascent GX10Jetson T5000 (Thor)~2Simulate output on NVIDIA Jetson AGX Thor Developer Kit192+ GB
fits · 71.6 GB free · up to 64K ctx
6 GPUs · 1 Mac class · 2 device families
B200~47Simulate output on B200B300~47Simulate output on B300Instinct MI350X~44Simulate output on Instinct MI350XM3 Ultra*256–512 GB~5Simulate output on Mac Studio (M3 Ultra, 256GB)8x H100 SXM5~159Simulate output on NVIDIA DGX H1008x A100 SXM4~97Simulate output on NVIDIA DGX A100 640GB
By VRAM tier · typical GPUs and Macs at Q4_K_M
Consumer GPUs (up to 32 GB) and Macs up to 64 GB: too small for this model (needs ≥ 88.9 GB)8 GB tier →12 GB tier →16 GB tier →24 GB tier →Macs →
Alternatives to GLM 5.3 Flash MXFP4
GLM 5 family · 6 models →Similar models
2| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Qwen3.8 Flash NextAlibaba | — | 109 GB | —needs > 24 GB | same size |
| BloomBigScience | — | 122.4 GB | —needs > 24 GB | same size |
Specifications
- Publisher
- OneNexus
- Family
- GLM 5
- Parameters
- 180.0B
- Architecture
- Glm5NextForConditionalGeneration
- Context Length
- 1,048,576 tokens
- Vocabulary Size
- 154,880
- Release Date
- 2026-08-30
- License
- MIT
Frequently Asked Questions
How much VRAM does GLM 5.3 Flash MXFP4 need?
GLM 5.3 Flash MXFP4 requires 109.8 GB of VRAM at Q4_K_M, or 361.8 GB at BF16. Full 1049K context adds up to 771.6 GB (881.4 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 180.0B × 4.8 bits ÷ 8 = 108 GB
KV Cache + Overhead ≈ 1.8 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.
KV Cache + Overhead ≈ 773.4 GB (at full 1049K context)
VRAM usage by quantization
Can NVIDIA GeForce RTX 5090 run GLM 5.3 Flash MXFP4?
No — GLM 5.3 Flash MXFP4 requires at least 78.3 GB at Q2_K, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
What's the best quantization for GLM 5.3 Flash MXFP4?
For GLM 5.3 Flash MXFP4, Q4_K_M (109.8 GB) offers the best balance of quality and VRAM usage. Q5_K_M (130.1 GB) provides better quality if you have the VRAM. The smallest option is Q2_K at 78.3 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run GLM 5.3 Flash MXFP4 on a Mac?
Yes — Mac Studio (M3 Ultra, 96GB) and 20 other Macs can run GLM 5.3 Flash MXFP4, the smallest at Q2_K. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 109.8 GB of usable unified memory (RAM minus macOS overhead), from Mac Pro M2 Ultra (128 GB).
Can I run GLM 5.3 Flash MXFP4 locally?
Yes — GLM 5.3 Flash MXFP4 can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is GLM 5.3 Flash MXFP4?
At Q4_K_M, GLM 5.3 Flash MXFP4 can reach ~44 tok/s on AMD Instinct MI350X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 109.8 × 0.65 = ~47 tok/s
Estimated speed at Q4_K_M (109.8 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of GLM 5.3 Flash MXFP4?
At Q4_K_M, the download is about 108.00 GB. The full-precision BF16 version is 359.98 GB. The smallest option (Q2_K) is 76.50 GB.
Which GPUs can run GLM 5.3 Flash MXFP4?
No single consumer GPU has enough VRAM to run GLM 5.3 Flash MXFP4 at Q4_K_M (109.8 GB). Multi-GPU or professional hardware is required.
Which devices can run GLM 5.3 Flash MXFP4?
18 devices with unified memory can run GLM 5.3 Flash MXFP4 at Q4_K_M (109.8 GB), including ASUS Ascent GX10, Mac Pro M2 Ultra (128 GB), Mac Pro M2 Ultra (192 GB), Mac Studio (M3 Ultra, 256GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.