ZE

ED Zephyria 48b i1

ZephyriaGGUF by mradermacherUpdated 2025-03-03
Download on Hugging Face
Parameters48.4B
Context128K
Layers123
Architecturellama
Smallest quant10.4 GB
Quick answer

ED Zephyria 48b i1 needs about 33.6 GB of memory at Q4_K_M (a 29.0 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 21 quants, from 10.4 GB (IQ1_S) to 39.7 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 16.5 GB.

ED Zephyria 48b i1 – 48.4B llama architecture model

ED Zephyria 48b i1 is a 48.4B parameter llama model with a 131072 token context window. It is quantized by mradermacher and available in multiple GGUF formats. The model is derived from SteelStorage/ED-Zephyria-48b.

Good at

  • Large context reasoning
  • High‑parameter inference
  • Fine‑grained text generation

Limits

  • No vision support
  • License not specified
  • Requires substantial VRAM for higher quants

Best for

  • Long‑form content creation
  • Complex dialogue systems
  • Research prototyping

Who it’s for. Developers and researchers needing a large llama model with extended context. Suitable for local deployment using llama.cpp with sufficient GPU memory.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S10.4 GB 15.0 GB
Too big Get file →
IQ1_M11.3 GB 15.9 GB
Too big Get file →
IQ2_XXS12.9 GB 17.5 GB
Too big Get file →
IQ2_XS14.4 GB 19.0 GB
Too big Get file →
IQ2_S15.2 GB 19.8 GB
Too big Get file →
IQ2_M16.5 GB 21.1 GB
Too big Get file →
Q2_K17.9 GB 22.5 GB
Too big Get file →
IQ3_XXS18.6 GB 23.2 GB
Too big Get file →
IQ3_XS19.9 GB 24.5 GB
Too big Get file →
Q3_K_S20.9 GB 25.5 GB
Too big Get file →
IQ3_S21.0 GB 25.6 GB
Too big Get file →
IQ3_M21.8 GB 26.4 GB
Too big Get file →
Q3_K_M23.3 GB 28.0 GB
Too big Get file →
Q3_K_L25.5 GB 30.1 GB
Too big Get file →
IQ4_XS25.9 GB 30.5 GB
Too big Get file →
Q4_027.4 GB 32.0 GB
Too big Get file →
Q4_K_S27.5 GB 32.1 GB
Too big Get file →
Q4_K_M29.0 GB 33.6 GB
Too big Get file →
Q5_K_S33.3 GB 37.9 GB
Too big Get file →
Q5_K_M34.2 GB 38.8 GB
Too big Get file →
Q6_K39.7 GB 44.3 GB
Too big Get file →

For 8 GB VRAM, use IQ1_S (10.36 GB file, 15 GB memory) or IQ1_M (11.32 GB file, 15.9 GB memory). For 12 GB VRAM, IQ2_XXS (12.92 GB file, 17.5 GB memory) or IQ2_XS (14.35 GB file, 19 GB memory) are suitable. Higher quants require more VRAM.

What GPU do you need for ED Zephyria 48b i1?

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: the best quant that fits is IQ1_M (11.3 GB file, about 15.9 GB in use).
  • 24 GB VRAM: the best quant that fits is IQ3_XXS (18.6 GB file, about 23.2 GB in use).
  • 32 GB VRAM: the best quant that fits is Q4_0 (27.4 GB file, about 32.0 GB in use).
  • 48 GB VRAM: the best quant that fits is Q6_K (39.7 GB file, about 44.3 GB in use).

Which ED Zephyria 48b i1 quant should you choose?

Q4_K_M is the usual starting point for ED Zephyria 48b i1: it balances file size and quality well. Here it is a 29.0 GB file.

If memory is tight, the smallest file is IQ1_S at 10.4 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters48.4B
Layers123
Trained context131,072 tokens
KV cache (fp16)492 KB per token

Source

Quantized bymradermacher
FamilyZephyria
Downloads (HF)1,374
Last updated2025-03-03

Frequently asked questions about ED Zephyria 48b i1

How much VRAM does ED Zephyria 48b i1 need?

About 33.6 GB at Q4_K_M with an 8K context window. That is the 29.0 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 ED Zephyria 48b i1 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 15.0 GB, so you would need a larger GPU or CPU offloading.

What is the file size of ED Zephyria 48b i1 GGUF?

It depends on the quant: IQ1_S is 10.4 GB, IQ1_M is 11.3 GB, IQ2_XXS is 12.9 GB, IQ2_XS is 14.4 GB, IQ2_S is 15.2 GB, IQ2_M is 16.5 GB and 15 more.

What context length does ED Zephyria 48b i1 support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 4.1 GB at 8K and 16.5 GB at 32K.

Where can I download ED Zephyria 48b i1 GGUF?

The files are on Hugging Face in the mradermacher/ED-Zephyria-48b-i1-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Can I run ED Zephyria 48b i1 on 8 GB VRAM?

Use the IQ1_S or IQ1_M quant, which estimate 15–15.9 GB memory at 8 k context.

Does this model support vision tasks?

No, the model has no vision capabilities.

What is the context length of this model?

The model supports 131072 tokens of context.

Other GGUFs of ED-Zephyria-48b

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Zephyria models

All Zephyria →

Other large models

All large models →