What are the hardware requirements for Mistral 7B MoEified 8x?

Mistral 7B MoEified 8x requires approximately 14.6 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 Mistral 7B MoEified 8x and how large is it?

Mistral 7B MoEified 8x is developed by kalomaze, a model with 7b 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 Mistral 7B MoEified 8x?

Mistral 7B MoEified 8x is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Mistral 7B MoEified 8x by kalomaze — VRAM 14.6GB, 32K context

Name: Mistral 7B MoEified 8x
Author: kalomaze

Mistral 7B MoEified 8x is an open-source language model by kalomaze. Features: 7b LLM, VRAM: 14.6GB, Context: 32K, License: apache-2.0, Quantized, LLM Explorer Score: 0.13.

Arxiv:2303.01610 Deploy:azure Endpoints compatible F16 Ggml Gguf Mixtral Quantized Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/kalomaze/Mistral-7b-MoEified-8x

Mistral 7B MoEified 8x Parameters and Internals

Model Type

dense language model

Use Cases

Areas:

Research

Primary Use Cases:

Adaptive computation for efficient token prediction

Additional Notes

The method involves modifying the dense language model by dividing MLP layers into "experts" and initializing router layers for unbiased expert activation.

Training Details

Methodology:

Expert layers division and router layers initialization for equal expert usage.

Model Architecture:

Slicing individual MLP layers into multiple experts with router layers initialized to ensure equal expert usage.

LLM Name	Mistral 7B MoEified 8x
Repository 🤗	https://huggingface.co/kalomaze/Mistral-7b-MoEified-8x
Model Size	7b
Required VRAM	14.6 GB
Updated	2026-07-18
Maintainer	kalomaze
Model Type	mixtral
Model Files	14.5 GB 1.0 GB: 1-of-15 1.0 GB: 2-of-15 1.0 GB: 3-of-15 1.0 GB: 4-of-15 1.0 GB: 5-of-15 1.0 GB: 6-of-15 1.0 GB: 7-of-15 1.0 GB: 8-of-15 1.0 GB: 9-of-15 1.0 GB: 10-of-15 1.0 GB: 11-of-15 1.0 GB: 12-of-15 1.0 GB: 13-of-15 1.0 GB: 14-of-15 0.6 GB: 15-of-15
GGML Quantization	Yes
GGUF Quantization	Yes
Quantization Type	ggml\|gguf
Model Architecture	MixtralForCausalLM
License	apache-2.0
Context Length	32768
Model Max Length	32768
Transformers Version	4.41.1
Tokenizer Class	PreTrainedTokenizerFast
Vocabulary Size	32000
Torch Data Type	bfloat16

Rank the Mistral 7B MoEified 8x 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 Mistral 7B MoEified 8x

Best Alternatives	Context / RAM	Downloads	Likes
Mixtral 7Bx2 MoE GGUF	32K / 4.8 GB	273	3
Laserxtral GGUF	32K / 8.8 GB	223	20
Laserxtral GGUF	32K / 8.8 GB	76	19
...izardLM 2 4x7B MoE EXL2 3 0bpw	32K / 9.4 GB	2	1
Buttercup 4x7B 6bpw EXL2	32K / 18.4 GB	6	2
BurningBruce 003 EXL2 B8.0	32K / 24.3 GB	7	1
...ixtral 2x7b DPO 8.0bpw H8 EXL2	32K / 13 GB	4	4
...ixtral 2x7b DPO 6.0bpw H6 EXL2	32K / 9.8 GB	6	1
...ixtral 2x7b DPO 4.0bpw H6 EXL2	32K / 6.7 GB	2	1
...ixtral 2x7b DPO 5.0bpw H6 EXL2	32K / 8.3 GB	1	1