LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  Agents   Benchmarks   LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

Mixtral 8x22B Instruct V0.1 FP8 V2 by comaniac

 »  All LLMs  »  comaniac  »  Mixtral 8x22B Instruct V0.1 FP8 V2   URL Share it on

  5   

Mixtral 8x22B Instruct V0.1 FP8 V2 is an open-source language model by comaniac. Features: 140.6b LLM, VRAM: 140.9GB, Context: 64K, MoE, Instruction-Based, LLM Explorer Score: 0.14.

  Conversational   Endpoints compatible   Fp8   Instruct   Mixtral   Moe   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/comaniac/Mixtral-8x22B-Instruct-v0.1-FP8-v2 

Mixtral 8x22B Instruct V0.1 FP8 V2 Benchmarks

LLME Score: 0.13705
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Mixtral 8x22B Instruct V0.1 FP8 V2 (comaniac/Mixtral-8x22B-Instruct-v0.1-FP8-v2)
🌟 Advertise your project 🚀

Mixtral 8x22B Instruct V0.1 FP8 V2 Parameters and Internals

Training Details 
Data Sources:
Ultrachat (mgoin/ultrachat_2k)
Data Volume:
2048 samples
Methodology:
Quantization with AutoFP8 with updated activation scaling factor names.
Context Length:
8192
Model Architecture:
Weights and activations are per-tensor quantized to float8_e4m3.
LLM NameMixtral 8x22B Instruct V0.1 FP8 V2
Repository 🤗https://huggingface.co/comaniac/Mixtral-8x22B-Instruct-v0.1-FP8-v2 
Model Size140.6b
Required VRAM140.9 GB
Updated2026-04-24
Maintainercomaniac
Model Typemixtral
Instruction-BasedYes
Model Files  4.9 GB: 1-of-29   4.9 GB: 2-of-29   4.9 GB: 3-of-29   4.9 GB: 4-of-29   4.9 GB: 5-of-29   4.9 GB: 6-of-29   4.9 GB: 7-of-29   4.9 GB: 8-of-29   4.9 GB: 9-of-29   4.9 GB: 10-of-29   4.9 GB: 11-of-29   4.9 GB: 12-of-29   4.9 GB: 13-of-29   4.9 GB: 14-of-29   4.9 GB: 15-of-29   4.9 GB: 16-of-29   4.9 GB: 17-of-29   4.9 GB: 18-of-29   4.9 GB: 19-of-29   4.9 GB: 20-of-29   5.0 GB: 21-of-29   5.0 GB: 22-of-29   5.0 GB: 23-of-29   4.9 GB: 24-of-29   4.9 GB: 25-of-29   4.9 GB: 26-of-29   4.9 GB: 27-of-29   4.9 GB: 28-of-29   3.4 GB: 29-of-29
Model ArchitectureMixtralForCausalLM
Context Length65536
Model Max Length65536
Transformers Version4.40.2
Tokenizer ClassLlamaTokenizer
Vocabulary Size32768
Torch Data Typebfloat16

Best Alternatives to Mixtral 8x22B Instruct V0.1 FP8 V2

Best Alternatives
Context / RAM
Downloads
Likes
Mixtral 8x22B Instruct V0.164K / 221.4 GB27071748
...ixtral 8x22B Instruct V0.1 FP864K / 140.9 GB5480
...igHuggyD Grey WizardLM 2 8x22B64K / 216.6 GB64
...x22B Instruct V0.1 FP8 Dynamic64K / 140.9 GB220
WizardLM 2 8x22B Beige64K / 221.4 GB34
...ral 8x22B Instruct V0.1 FP8 V164K / 140.9 GB80
Mixtral 8x22B GO Instruct V164K / 280.8 GB52
Mixtral 8x22B Instruct V0.164K / 221.4 GB171
Mixtral 8x22b Instruct Oh64K / 221.6 GB1029
... 8x22B V0.1 Instruct Sft En De64K / 212 GB121
Note: green Score (e.g. "73.2") means that the model is better than comaniac/Mixtral-8x22B-Instruct-v0.1-FP8-v2.

Rank the Mixtral 8x22B Instruct V0.1 FP8 V2 Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

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  

What open-source LLMs or SLMs are you in search of? 53254 in total.

Email us: info@extractum.io. Our Privacy Policy  | Terms and Conditions  | Suggest an improvement.
Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a
  Support LLM Explorer