LLM EXPLORER 59,516 MODELS INDEXED

Midnight Miqu 103B V1.0 EXL2 4.25bpw by Dracones

By Dracones · 3 downloads
HF Repo ↗ Copy URL

Midnight Miqu 103B V1.0 EXL2 4.25bpw is an open-source language model by Dracones. Features: 103b LLM, VRAM: 55.3GB, Context: 31K, Quantized, LLM Explorer Score: 0.11.

  Conversational   Endpoints compatible   Exl2   Llama   Merge   Mergekit   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/Dracones/Midnight-Miqu-103B-v1.0_exl2_4.25bpw 

Midnight Miqu 103B V1.0 EXL2 4.25bpw Parameters and Internals

Additional Notes 
Details about the model and the merge info can be found at the above model page.
LLM NameMidnight Miqu 103B V1.0 EXL2 4.25bpw
Repository 🤗https://huggingface.co/Dracones/Midnight-Miqu-103B-v1.0_exl2_4.25bpw 
Model Size103b
Required VRAM55.3 GB
Updated2026-07-31
MaintainerDracones
Model Typellama
Model Files  8.6 GB: 1-of-7   8.6 GB: 2-of-7   8.5 GB: 3-of-7   8.6 GB: 4-of-7   8.5 GB: 5-of-7   8.5 GB: 6-of-7   4.0 GB: 7-of-7
Quantization Typeexl2
Model ArchitectureLlamaForCausalLM
Context Length32764
Model Max Length32764
Transformers Version4.36.2
Tokenizer ClassLlamaTokenizer
Padding Token<unk>
Vocabulary Size32000
Torch Data Typefloat16

Best Alternatives to Midnight Miqu 103B V1.0 EXL2 4.25bpw

Best Alternatives
Context / RAM
Downloads
Likes
... Dark Miqu 103B 5.0bpw H8 EXL231K / 65.1 GB100
...qu 103B V1.5 EXL2 5.0bpw Rpcal31K / 64.9 GB91
... Miqu 103B V1.5 3.5bpw H6 EXL231K / 45.6 GB21
...qu 103B V1.5 EXL2 4.0bpw Rpcal31K / 52.1 GB31
...qu 103B V1.5 EXL2 3.0bpw Rpcal31K / 39.3 GB21
...ght Miqu 103B V1.0 EXL2 4.5bpw31K / 58.7 GB71
...ght Miqu 103B V1.0 EXL2 4.0bpw31K / 52.2 GB61
...ght Miqu 103B V1.0 EXL2 3.5bpw31K / 45.7 GB91
...ght Miqu 103B V1.0 EXL2 3.0bpw31K / 39.2 GB42
...ght Miqu 103B V1.0 EXL2 5.0bpw31K / 64.9 GB51
Note: green Score (e.g. "73.2") means that the model is better than Dracones/Midnight-Miqu-103B-v1.0_exl2_4.25bpw.