gemma-4-31B-it needs about 34 GB of GPU memory for FP8 weights. The smallest fitting setup at 32k context is 1× NVIDIA B200. It supports 64 concurrent conversations with SGLang defaults.
GPU requirements by context length
8k context
GPU
GPUs
Concurrent users
NVIDIA B200
1
64
NVIDIA H200 SXM
1
64
NVIDIA H100 SXM
1
48
NVIDIA H100 NVL
1
64
NVIDIA A100 80GB SXM
1
42
NVIDIA L40S
1
6
NVIDIA RTX PRO 6000 Blackwell Server Edition
1
64
NVIDIA H20
1
64
32k context
GPU
GPUs
Concurrent users
NVIDIA B200
1
64
NVIDIA H200 SXM
1
47
NVIDIA H100 SXM
1
20
NVIDIA H100 NVL
1
28
NVIDIA A100 80GB SXM
1
18
NVIDIA L40S
1
2
NVIDIA RTX PRO 6000 Blackwell Server Edition
1
29
NVIDIA H20
1
29
128k context
GPU
GPUs
Concurrent users
NVIDIA B200
1
21
NVIDIA H200 SXM
1
14
NVIDIA H100 SXM
1
6
NVIDIA H100 NVL
1
8
NVIDIA A100 80GB SXM
1
5
NVIDIA RTX PRO 6000 Blackwell Server Edition
1
8
NVIDIA H20
1
8
NVIDIA L40S
2
7
FP8 weights, SGLang defaults, estimates.
Weights by precision
Weight memory
Precision
Weights
Smallest fitting setup at 32k
BF16
about 63 GB
1x NVIDIA B200
FP8
about 34 GB
1x NVIDIA B200
INT4 (AWQ)
about 20 GB
1x NVIDIA B200
Model notes
Attention
Hybrid attention, full + sliding-window layers
Architecture
Dense, 32.7B parameters
Sliding window
1,024 tokens on 50 of 60 layers
Layers
60
Hidden size
5,376
Vocabulary
262,144
The sliding-window design caps this model's memory appetite, past 1,024 tokens, 50 of its 60 layers stop charging for longer conversations.
Frequently asked questions
Will google/gemma-4-31B-it run on a single H100?
Yes, google/gemma-4-31B-it fits on a single H100 SXM at 32k context with FP8 weights, supporting 20 concurrent conversations.
What is the cheapest GPU setup for google/gemma-4-31B-it?
The cheapest fitting setup at 32k context is 1x NVIDIA B200 with FP8 weights and SGLang defaults.
How much VRAM does google/gemma-4-31B-it need at 128k context?
At 128k context on the cheapest fitting setup, google/gemma-4-31B-it uses about 34 GB of VRAM per GPU for FP8 weights and about 5.8 GB per conversation for the KV cache. The total VRAM needed depends on the GPU count and parallelism configuration.
Can I run google/gemma-4-31B-it with INT4 quantization?
Yes, google/gemma-4-31B-it has INT4 weight figures. Its weights take about 20 GB in INT4 versus about 34 GB in FP8. INT4 can make the model fit on fewer GPUs, but check inference quality for your workload.
Want a different context length, GPU, or quantization? The calculator runs the same numbers live, preloaded with this model.