What are the hardware requirements for ChaoticSoliloquy 4x8B 8bpw H8 EXL2?

ChaoticSoliloquy 4x8B 8bpw H8 EXL2 requires approximately 25.3 GB of VRAM and supports a context window of 8K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed ChaoticSoliloquy 4x8B 8bpw H8 EXL2 and how large is it?

ChaoticSoliloquy 4x8B 8bpw H8 EXL2 is developed by JayhC, a model with 8b 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 ChaoticSoliloquy 4x8B 8bpw H8 EXL2?

ChaoticSoliloquy 4x8B 8bpw H8 EXL2 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

ChaoticSoliloquy 4x8B 8bpw H8 EXL2 by JayhC — VRAM 25.3GB, 8K context

Name: ChaoticSoliloquy 4x8B 8bpw H8 EXL2
Author: JayhC

ChaoticSoliloquy 4x8B 8bpw H8 EXL2 is an open-source language model by JayhC. Features: 8b LLM, VRAM: 25.3GB, Context: 8K, License: llama3, MoE, Quantized, LLM Explorer Score: 0.12.

8-bit Conversational En Endpoints compatible Exl2 Mixtral Moe Quantized Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/JayhC/ChaoticSoliloquy-4x8B-8bpw-h8-exl2

ChaoticSoliloquy 4x8B 8bpw H8 EXL2 Parameters and Internals

Model Type

moe

Additional Notes

The model is designed to outperform Mixtral 8x7B in RP/ERP tasks using an MoE architecture.

Supported Languages

en (full proficiency)

Training Details

Data Sources:

ChaoticNeutrals_Poppy_Porpoise-v0.6-L3-8B, jeiku_Chaos_RP_l3_8B, openlynn_Llama-3-Soliloquy-8B, Sao10K_L3-Solana-8B-v1

Methodology:

exl2 quantization using default exllamav2 calibration dataset

Model Architecture:

MoE with random gate mode, bfloat16, 2 experts per token

Input Output

Input Format:

Llama 3

LLM Name	ChaoticSoliloquy 4x8B 8bpw H8 EXL2
Repository 🤗	https://huggingface.co/JayhC/ChaoticSoliloquy-4x8B-8bpw-h8-exl2
Model Size	8b
Required VRAM	25.3 GB
Updated	2026-07-21
Maintainer	JayhC
Model Type	mixtral
Model Files	8.5 GB: 1-of-3 8.5 GB: 2-of-3 8.3 GB: 3-of-3
Supported Languages	en
Quantization Type	exl2
Model Architecture	MixtralForCausalLM
License	llama3
Context Length	8192
Model Max Length	8192
Transformers Version	4.37.2
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|begin_of_text\|>
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the ChaoticSoliloquy 4x8B 8bpw H8 EXL2 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 ChaoticSoliloquy 4x8B 8bpw H8 EXL2

Best Alternatives	Context / RAM	Downloads	Likes
...Storm V1.15 4x8B B 8 0bpw EXL2	8K / 25.2 GB	6	0
... SnowStorm 4x8B 6.5bpw H8 EXL2	8K / 21 GB	5	2
L3 SnowStorm 4x8B 8bpw H8 EXL2	8K / 25.2 GB	6	1
...oquy V2 4x8B Test 8bpw H8 EXL2	8K / 24.9 GB	5	1
...liloquy V1.5 4x8B 8bpw H8 EXL2	8K / 25.3 GB	10	2
...loquy V1.5 4x8B 6.5bpw H8 EXL2	8K / 21 GB	6	2
...oquy 4x8B Bpw600 H6 EXL2 Rpcal	8K / 19.4 GB	5	2
...oquy 4x8B Bpw500 H6 EXL2 Rpcal	8K / 16.4 GB	4	1
...oquy 4x8B Bpw800 H8 EXL2 Rpcal	8K / 25.3 GB	0	1
...ticSoliloquy 4x8B 6bpw H6 EXL2	8K / 19.4 GB	2	3

Note: green Score (e.g. "73.2") means that the model is better than JayhC/ChaoticSoliloquy-4x8B-8bpw-h8-exl2.