LLM EXPLORER 59,420 MODELS INDEXED

CognitiveFusion 4x7b MoE EXL2 3.5BPW by SytanSD

By SytanSD · 8 downloads
HF Repo ↗ Copy URL

CognitiveFusion 4x7b MoE EXL2 3.5BPW is an open-source language model by SytanSD. Features: 7b LLM, VRAM: 10.8GB, Context: 8K, License: apache-2.0, MoE, Quantized, LLM Explorer Score: 0.11.

  En   Endpoints compatible   Exl2   Mixtral   Moe   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/SytanSD/CognitiveFusion_4x7b_MoE_Exl2_3.5BPW 

CognitiveFusion 4x7b MoE EXL2 3.5BPW Parameters and Internals

Additional Notes 
The model is available in different BPW quants, including 8.0BPW and 3.5BPW, with plans for additional quants.
Supported Languages 
en (native)
LLM NameCognitiveFusion 4x7b MoE EXL2 3.5BPW
Repository 🤗https://huggingface.co/SytanSD/CognitiveFusion_4x7b_MoE_Exl2_3.5BPW 
Model Size7b
Required VRAM10.8 GB
Updated2026-08-08
MaintainerSytanSD
Model Typemixtral
Model Files  8.6 GB: 1-of-2   2.2 GB: 2-of-2
Supported Languagesen
Quantization Typeexl2
Model ArchitectureMixtralForCausalLM
Licenseapache-2.0
Context Length8192
Model Max Length8192
Transformers Version4.36.2
Tokenizer ClassLlamaTokenizer
Padding Token<s>
Vocabulary Size32000
Torch Data Typebfloat16

Best Alternatives to CognitiveFusion 4x7b MoE EXL2 3.5BPW

Best Alternatives
Context / RAM
Downloads
Likes
...izardLM 2 4x7B MoE EXL2 3 0bpw32K / 9.4 GB21
Buttercup 4x7B 6bpw EXL232K / 18.4 GB62
BurningBruce 003 EXL2 B8.032K / 24.3 GB71
...ixtral 2x7b DPO 8.0bpw H8 EXL232K / 13 GB44
...ixtral 2x7b DPO 6.0bpw H6 EXL232K / 9.8 GB61
...ixtral 2x7b DPO 4.0bpw H6 EXL232K / 6.7 GB21
...ixtral 2x7b DPO 5.0bpw H6 EXL232K / 8.3 GB11
...t V0.2 2x7B MoE 6.0bpw H6 EXL232K / 9.9 GB31
...lue Orchid 2x7b 8.0bpw H8 EXL28K / 13 GB75
...lue Orchid 2x7b 5.0bpw H6 EXL28K / 8.2 GB84
Note: green Score (e.g. "73.2") means that the model is better than SytanSD/CognitiveFusion_4x7b_MoE_Exl2_3.5BPW.