Hermes 4 70B FP8 — Hardware Requirements & GPU Compatibility
ChatReasoningRoleplay
Hermes 4 70B FP8 is a 70.6B-parameter open language model from Nous Research in the Hermes family. It supports a context window of up to 131,072 tokens. At BF16 it needs about 142.09 GB of VRAM — see which GPUs and Macs can run it below.
3.0K downloads
34 likesSpecifications
- Publisher
- Nous Research
- Family
- Hermes
- Parameters
- 70.6B
- Architecture
- LlamaForCausalLM
- Context Length
- 131,072 tokens
- Vocabulary Size
- 128,256
- Release Date
- 2025-09-12
- License
- Llama 3 Community
Get Started
HuggingFace
How Much VRAM Does Hermes 4 70B FP8 Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| BF16 | 16.00 | 142.1 GB | 184.4 GB | 141.12 GB | Brain floating point 16 — preferred for training |
Which GPUs Can Run Hermes 4 70B FP8?
BF16 · 142.1 GBHermes 4 70B FP8 (BF16) requires 142.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 185+ GB is recommended. Using the full 131K context window can add up to 42.3 GB, bringing total usage to 184.4 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run Hermes 4 70B FP8?
BF16 · 142.1 GB4 devices with unified memory can run Hermes 4 70B FP8, including NVIDIA DGX H100, NVIDIA DGX A100 640GB, Mac Pro M2 Ultra (192 GB).
Runs great
— Plenty of headroom
Decent
— Enough memory, may be tightRelated Models
Frequently Asked Questions
- How much VRAM does Hermes 4 70B FP8 need?
Hermes 4 70B FP8 requires 142.1 GB of VRAM at BF16. Full 131K context adds up to 42.3 GB (184.4 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 70.6B × 16 bits ÷ 8 = 141.1 GB
KV Cache + Overhead ≈ 1 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 43.3 GB (at full 131K context)
VRAM usage by quantization
BF16142.1 GBBF16 + full context184.4 GB- Can NVIDIA GeForce RTX 5090 run Hermes 4 70B FP8?
No — Hermes 4 70B FP8 requires at least 142.1 GB at BF16, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- Can I run Hermes 4 70B FP8 on a Mac?
Hermes 4 70B FP8 requires at least 142.1 GB at BF16, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run Hermes 4 70B FP8 locally?
Yes — Hermes 4 70B FP8 can run locally on consumer hardware. At BF16 quantization it needs 142.1 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Hermes 4 70B FP8?
At BF16, Hermes 4 70B FP8 can reach ~21 tok/s on AMD Instinct MI300X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: AMD Instinct MI300X → 5300 ÷ 142.1 × 0.55 = ~21 tok/s
Estimated speed at BF16 (142.1 GB)
~21 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Hermes 4 70B FP8?
At BF16, the download is about 141.12 GB.
- Which GPUs can run Hermes 4 70B FP8?
No single consumer GPU has enough VRAM to run Hermes 4 70B FP8 at BF16 (142.1 GB). Multi-GPU or professional hardware is required.
- Which devices can run Hermes 4 70B FP8?
4 devices with unified memory can run Hermes 4 70B FP8 at BF16 (142.1 GB), including Mac Pro M2 Ultra (192 GB), Mac Studio M2 Ultra (192 GB), NVIDIA DGX A100 640GB, NVIDIA DGX H100. Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.