mar llama3.1 8B needs about 17.6 GB of memory at its main file (a 16.1 GB file) with an 8K context window, so it fits on a 24 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.
Hermes 3 Llama‑3.1 8B – a generalist language model with advanced roleplay and agentic capabilities
Hermes 3 is a 8B parameter Llama‑3.1 model optimized for roleplay, reasoning, and multi‑turn conversation. It supports ChatML prompts and offers improved function calling and structured output. The model is available in a single GGUF quantization.
Good at
- Advanced roleplay
- Multi‑turn conversation
- Structured output
Limits
- No vision support
- Requires 17.6 GB VRAM at 8 k context
- Single quantization available
Best for
- Chat assistants
- Role‑playing scenarios
- Code generation
Who it’s for. Developers and researchers needing a versatile LLM for conversational AI and agentic tasks. The model is suitable for applications that benefit from structured prompts and detailed roleplay.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 16.1 GB | 17.6 GB | Too big | Get file → |
The only available GGUF file is 16.07 GB with an estimated memory need of 17.6 GB at 8 k context. For systems with 8 GB VRAM, the model cannot run. For 16 GB VRAM, it is borderline; for 24 GB VRAM, it runs comfortably.
What GPU do you need for mar llama3.1 8B?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is the main file (16.1 GB file, about 17.6 GB in use).
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about mar llama3.1 8B
How much VRAM does mar llama3.1 8B need?
About 17.6 GB at its main file with an 8K context window. That is the 16.1 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run mar llama3.1 8B on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 17.6 GB, so you would need a larger GPU or CPU offloading.
What is the file size of mar llama3.1 8B GGUF?
It depends on the quant: file is 16.1 GB.
What context length does mar llama3.1 8B support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.
Where can I download mar llama3.1 8B GGUF?
The files are on Hugging Face in the xteapot/mar-llama3.1-8B repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does mar llama3.1 8B use?
Hugging Face lists the license as llama3. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run Hermes 3 on 8 GB VRAM?
No, the model requires approximately 17.6 GB VRAM at 8 k context.
What prompt format does Hermes 3 use?
Hermes 3 uses ChatML, the same format as OpenAI’s ChatGPT API.
Does Hermes 3 support vision tasks?
No, the model does not support vision capabilities.