Llama
How much VRAM does Llama 4 Maverick (17B active / 400B total) need?
Llama 4 Maverick (17B active / 400B total) has 400B parameters, so its weights alone take about 800 GB of VRAM at FP16 — or roughly 200 GB quantized to INT4. Real serving adds KV cache on top, which scales with your context length and concurrent requests. Size the exact figure for your workload below.
MoE: all 128 experts resident in memory; 1 active per token for compute.
Total params
400B
Active params
17B
Layers
48
Hidden size
5120
Attention heads
40
KV heads (GQA)
8
Vocab size
202,048
Native context window
1,048,576
Native precision
BF16
Experts (total)
128
Experts active / token
1
Size Llama 4 Maverick (17B active / 400B total) for your workload
Opens the free calculator with Llama 4 Maverick (17B active / 400B total) loaded at BF16 weights and FP16 KV cache. Set your context length and traffic to get exact GPU memory and a ranked list of cloud instances that fit — no signup.
Size this model →