How much VRAM does Hermes 3 405B Llama-3.1 need?
Hermes 3 405B Llama-3.1 needs about 240.9 GB of VRAM at Q4_K_M, or 421.5 GB at Q8_0 and 792.8 GB at FP16. No single GPU in our set holds it even at Q4_K_M; it needs multi-GPU or offload. Weights plus runtime overhead. Budget separately for KV cache, which at 128K is a material share.
Weights-plus-runtime footprint across the six most common quantizations. KV cache is context-dependent and comes on top. At 128K it is gigabytes on its own, before the weights; size it precisely in the GPU Memory Calculator.
Can your GPU run Hermes 3 405B Llama-3.1?
Pick your card — the answer below is computed by the same engine that produces every figure on this page.
H200 141GB cannot hold Hermes 3 405B Llama-3.1 at Q4_K_M: it needs 240.9 GB against 133.9 GB usable; a heavier quantization, a larger card, or splitting across GPUs are the ways forward.
Weights plus runtime overhead. KV cache is not included — it depends on context length and this model’s published architecture.
| Weights (Q4_K_M) | 228.7 GB |
| Runtime overhead | 12.2 GB |
| Estimated total | 240.9 GB |
| Usable VRAM (95% of 141 GB) | 133.9 GB |
Across every GPU we track
- RTX 3060 12GBNot enough-229.5 GB
- RTX 4060 Ti 16GBNot enough-225.7 GB
- RTX 4070 Ti SUPERNot enough-225.7 GB
- RTX 3090Not enough-218.1 GB
- RTX 4090Not enough-218.1 GB
- RTX 5090Not enough-210.5 GB
- A100 40GBNot enough-202.9 GB
- RTX A6000Not enough-195.3 GB
- L40SNot enough-195.3 GB
- A100 80GBNot enough-164.9 GB
- H100 80GBNot enough-164.9 GB
- H200 141GBNot enough-106.9 GB
Estimates, not guarantees: real usage moves with runtime, driver, batch size and context. Size a specific context window in the GPU Memory Calculator.
How was this number calculated? Every figure here comes from Bitpute’s documented calculation methodology — parameters, precision, quantization, runtime overhead and usable VRAM, each formula written out in full.
Memory by quantization
| Quant | Weights | + runtime overhead |
|---|---|---|
| FP16 | 754.4 GB | 792.8 GB |
| Q8_0 | 400.8 GB | 421.5 GB |
| Q6_K | 309.3 GB | 325.5 GB |
| Q5_K_M | 268.3 GB | 282.4 GB |
| Q4_K_M | 228.7 GB | 240.9 GB |
| Q4_0 | 212.2 GB | 223.5 GB |
Overhead = 0.75 GB + 5% of weights (CUDA context, buffers). Add KV cache on top. Weights dominate at this size, but a 128K window still costs gigabytes.
Which quantization should you actually run?
Hermes 3 405B Llama-3.1 is a multi-GPU or datacenter model: even Q4_K_M needs about 240.9 GB, beyond any single card. Plan for several 80 GB GPUs or a hosted endpoint; FP16 (792.8 GB) is server-only.
Single-GPU fit at Q4_K_M (240.9 GB + KV)
RTX 3060 12GBRTX 4060 Ti 16GBRTX 4070 Ti SUPERRTX 3090RTX 4090RTX 5090A100 40GBRTX A6000L40SA100 80GBH100 80GBH200 141GB
Common questions
How much VRAM does Hermes 3 405B Llama-3.1 need?
Hermes 3 405B Llama-3.1 needs about 240.9 GB of VRAM at Q4_K_M, 421.5 GB at Q8_0, or 792.8 GB at FP16. That is weights plus about 0.75 GB of runtime overhead and 5% of weight size; KV cache is additional and depends on context length.
What GPU can run Hermes 3 405B Llama-3.1?
No single GPU in our set fits it at Q4_K_M; it needs multiple GPUs or CPU offload. At FP16 no single GPU in our set is sufficient.
Can Hermes 3 405B Llama-3.1 run on 24 GB?
Not at Q4_K_M: it needs about 240.9 GB, more than the ~22.8 GB usable on a 24 GB card. Use a larger card, multiple GPUs, or offload.
More in the Hermes family
Hermes 2 Pro 8B Llama-3Hermes 3 8B Llama-3.1Hermes 3 70B Llama-3.1