What are the hardware requirements for Zion Alpha?

Zion Alpha requires approximately 14.4 GB of VRAM and supports a context window of 32K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Zion Alpha and how large is it?

Zion Alpha is developed by SicariusSicariiStuff, a model with 7.2b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Zion Alpha?

Zion Alpha is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Zion Alpha by SicariusSicariiStuff — VRAM 14.4GB, 32K context

Name: Zion Alpha
Author: SicariusSicariiStuff

Zion Alpha is an open-source language model by SicariusSicariiStuff. Features: 7.2b LLM, VRAM: 14.4GB, Context: 32K, License: apache-2.0, Quantized, LLM Explorer Score: 0.18.

En Endpoints compatible Gguf He Mistral Q3 Quantized Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/SicariusSicariiStuff/Zion_Alpha

Zion Alpha Parameters and Internals

Model Type

translation, comprehension

Use Cases

Areas:

Research, Commercial applications

Applications:

Translation, Comprehension

Additional Notes

Future plans include SLERP merging and potentially creating larger models using MergeKit.

Supported Languages

Hebrew (understands and comprehends), English (decents translation)

Training Details

Methodology:

used SOTA techniques and personal insights

Model Architecture:

Based on Mistral 7B, tokenizer not altered.

Input Output

Input Format:

Debug-deterministic with specified parameters.

Accepted Modalities:

text

Performance Tips:

No specific tips provided beyond using the recommended prompt settings.

Release Notes

Version:

v0.1

Date:

2024-06-04

Notes:

Acquired the highest SNLI score among open source models in Hebrew.

LLM Name	Zion Alpha
Repository 🤗	https://huggingface.co/SicariusSicariiStuff/Zion_Alpha
Model Size	7.2b
Required VRAM	14.4 GB
Updated	2026-08-04
Maintainer	SicariusSicariiStuff
Model Type	mistral
Model Files	3.5 GB 4.9 GB: 1-of-3 5.0 GB: 2-of-3 4.5 GB: 3-of-3
Supported Languages	en he
GGUF Quantization	Yes
Quantization Type	q3\|gguf
Model Architecture	MistralForCausalLM
License	apache-2.0
Context Length	32768
Model Max Length	32768
Transformers Version	4.36.2
Tokenizer Class	LlamaTokenizer
Padding Token	</s>
Vocabulary Size	32000
Torch Data Type	bfloat16

Rank the Zion Alpha capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Zion Alpha

Best Alternatives	Context / RAM	Downloads	Likes
Spydaz Web AGI DeepThink	128K / 14.4 GB	14	1
...daz Web AI Pre Train Align 004	128K / 14.4 GB	28	0
LCARS AI 001	32K / 14.4 GB	14	4
Mixtral Quantized	32K / 14.4 GB	6	0
Mayonnaise 4in1 02	32K / 14.4 GB	105	0
Zsql En Postgres	32K / 14.4 GB	64	5
MHENNlitv3	32K / 28.9 GB	12	0
MHENNlitv2	32K / 28.9 GB	17	0
MHENNcodemathv2	32K / 28.9 GB	10	0
MHENNcodemath	32K / 28.9 GB	17	0