Z.ai GLM-4.5

GLM-4.5 hardware requirements

Needs at least 4 GPUs at BF16 (Instinct MI350X OAM). For 1,000 tokens/s at 32k context you need 16 x H100 SXM5 80GB at $28.64/hr or 16 x RTX PRO 6000 Blackwell Max-Q Workstation Edition at $8.00/hr.

Parameters

358B

Active per token

32.0B

mixture of experts

KV cache per token

368 KB

16-bit cache

Native precision

BF16

Rent the cheapest fit: 8 x RTX PRO 6000 Blackwell Max-Q Workstation Edition from $4.00/hr Model card

Bare minimum

Fewest GPUs

4 x Instinct MI350X OAM

BF16, 4,096 tokens, one stream

Tokens/s: 330
Per hour: $24.64
Per million tokens: $20.77
Memory used: 65%

Rent on DigitalOcean

Cheapest per hour

8 x RTX PRO 6000 Blackwell Max-Q Workstation Edition

BF16, 4,096 tokens, one stream, live price

Tokens/s: 139
Per hour: $4.00
Per million tokens: $7.97
Memory used: 98%

Rent on RunPod

Cheapest per million tokens

16 x RTX PRO 6000 Blackwell Max-Q Workstation Edition

BF16, 32,768 tokens, 32 streams

Tokens/s: 1.1k
Per hour: $8.00
Per million tokens: $1.98
Memory used: 76%

Rent on RunPod

Configurations

Every datacenter GPU in the catalog, sized for these settings. Change precision, context and concurrency and the whole page follows. Each point is one replica: the smallest tensor-parallel group of that GPU that holds the model.

Precision Context Streams

Vendor

Priced only

Priced configurations only; 11 more fit without a live price and appear in the table. Marker size grows with the tensor-parallel width. Roofline estimates.

GPU	TP	Memory used	Tokens/s	Per stream	Per hour	Per M tokens	kW
Instinct MI455X OAM	4	67%	~4.4k	137				Specs
Rubin SXM	8	50%	7.8k	244				Specs
Instinct MI355X OAM	8	50%	2.8k	88.9			11.2	Specs
GB300 NVL72 GPU 288GB	8	50%	2.8k	88.9	$68.96	$6.74		Rent
GB200 NVL72 GPU 186GB	8	78%	2.8k	88.9	$84.00	$8.21		Rent
Instinct MI350X OAM	8	50%	2.8k	88.9	$49.28	$4.81	8.0	Rent
B300 SXM 262GB	8	55%	2.8k	88.9	$59.20	$5.78	11.2	Rent
B200 SXM 180GB	8	80%	2.7k	85.5	$32.72	$3.32	8.0	Rent
B100 SXM 192GB	8	75%	2.8k	88.9			5.6	Specs
Instinct MI325X OAM	8	56%	2.1k	66.6	$30.40	$3.96	8.0	Rent
Instinct MI300X 192GB	8	75%	1.9k	58.9	$19.12	$2.82	6.0	Rent
H200 SXM 141GB 2 nodes	16	51%	3.0k	94.1	$47.84	$4.41	11.2	Rent
H100 SXM5 80GB 2 nodes	16	91%	2.1k	65.7	$28.64	$3.79	11.2	Rent
Data Center GPU Max 1550 128GB 2 nodes	16	57%	2.1k	64.2			9.6	Specs
H200 NVL 141GB 2 nodes	16	51%	~3.0k	94.1	$60.64	$5.59	9.6	Rent
H100 NVL 94GB 2 nodes	16	77%	~2.4k	76.4	$51.04	$5.80	6.4	Rent
H100 PCIe 80GB 2 nodes	16	91%	1.3k	39.2	$40.00	$8.86	5.6	Rent
H800 PCIe 80GB 2 nodes	16	91%	1.3k	39.2			5.6	Specs
RTX PRO 6000 Blackwell Workstation Edition 2 nodes	16	76%	1.1k	35.1	$28.80	$7.12	9.6	Rent
RTX PRO 6000 Blackwell Server Edition 2 nodes	16	76%	~1.0k	31.3	$9.44	$2.62	9.6	Rent
RTX PRO 6000 Blackwell Max-Q Workstation Edition 2 nodes	16	76%	1.1k	35.1	$8.00	$1.98	4.8	Rent
MI250X 128GB 2 nodes	16	57%	2.1k	64.2			9.0	Specs
Instinct MI250 128GB 2 nodes	16	57%	2.1k	64.2			8.0	Specs
A800 PCIe 80GB 2 nodes	16	91%	1.2k	37.9			4.8	Specs
A100 PCIe 80GB 2 nodes	16	91%	1.2k	37.9	$21.60	$4.94	4.8	Rent
A100 SXM4 80GB 2 nodes	16	91%	1.3k	40.0	$22.40	$4.86	6.4	Rent
H20 141GB HBM3e 2 nodes	16	51%	2.5k	78.4			6.4	Specs
H20 96GB 2 nodes	16	76%	2.5k	78.4			6.4	Specs

AMD vs NVIDIA

At BF16 the NVIDIA pick is 16 x RTX PRO 6000 Blackwell Max-Q Workstation Edition at about 1.1k tokens/s for $8.00/hr; the AMD pick is 8 x Instinct MI300X 192GB at about 1.9k tokens/s for $19.12/hr. Per rental dollar NVIDIA delivers 1.43x the tokens of AMD here.

NVIDIA 20 parts fit

Best value: 16 x RTX PRO 6000 Blackwell Max-Q Workstation Edition, 1.1k tok/s for $8.00/hr

Fewest GPUs: 8 x Rubin SXM, 7.8k tok/s

AMD 7 parts fit

Best value: 8 x Instinct MI300X 192GB, 1.9k tok/s for $19.12/hr

Fewest GPUs: 4 x Instinct MI455X OAM, 4.4k tok/s

Intel 1 parts fit

Fewest GPUs: 16 x Data Center GPU Max 1550 128GB, 2.1k tok/s

Fleet what-if

Compare whole fleets at these settings: a thousand of one part against a hundred of another. Each fleet splits into replicas of its tensor-parallel width; GPUs left over sit idle.

A GPUs Part

B GPUs Part

Fleet	Replicas	Idle	Tokens/s	Per hour	Power	Per M tokens
A 1,008 x H200 SXM 141GB	63 x TP16	0	190k	$3,014	706 kW	$4.41
B 104 x GB200 NVL72 GPU 186GB	13 x TP8	0	37k	$1,092	kW	$8.21

Size GLM-4.5 for a tokens per second target

The calculator starts from your traffic instead of a fleet: set a target rate and read the replica count for every GPU.

Open in the LLM GPU Calculator

Frequently asked questions

How much GPU memory does GLM-4.5 need?

The weights take 717 GB at the native BF16 precision (717 GB at BF16, 358 GB at FP8, 202 GB at INT4). Each concurrent stream adds 368 KB of KV cache per token: 1.5 GB at 4,096 tokens and 12.3 GB at 32,768 tokens.

What is the bare minimum to run GLM-4.5?

4 x Instinct MI350X OAM at BF16 with 4,096 tokens of context and one stream. The cheapest live rental for that footprint is 8 x RTX PRO 6000 Blackwell Max-Q Workstation Edition at $4.00 per hour.

How many H100s do you need to run GLM-4.5?

16 H100s at BF16 for 4,096 tokens of context and one stream. For 32,768 tokens and 32 concurrent streams the tensor-parallel group is 16 H100s, and 1,000 tokens per second takes 16 H100s across 1 replicas.

How much does GLM-4.5 cost per million tokens on an H100?

About $3.79 per million output tokens at BF16, 32,768 tokens of context and 32 streams, using the cheapest live on-demand price of $1.79 per GPU-hour. The cheapest part per token is 16 x RTX PRO 6000 Blackwell Max-Q Workstation Edition at $1.98.

What precision does GLM-4.5 ship in?

The published checkpoint is BF16. FP8 and INT4 figures describe post-training quantisations that halve and quarter the footprint at a small accuracy cost.

All figures are roofline estimates from the model's config.json and each GPU's published memory, bandwidth and tensor-core peak, with fixed efficiency factors. Prices are the cheapest live on-demand listing per GPU when the site has one. They are a planning floor; a tuned serving stack can do better. Active parameters per the Z.ai model card.