LLM EXPLORER 64,924 MODELS INDEXED

Swift1.5 Qwen3.8 Flash Next Q4 MTPLX by Ecmcfee

By Ecmcfee · 574 downloads
HF Repo ↗ Copy URL

Swift1.5 Qwen3.8 Flash Next Q4 MTPLX is an open-source language model by Ecmcfee. Features: 20.9b LLM, VRAM: 72.8GB, License: other, Quantized, LLM Explorer Score: 0.34.

  4-bit Base model:quantized:ukisai/sw... Base model:ukisai/swift1.5-qwe...   Conversational   Efficient-thinking   Image-text-to-text   M5   Mlx   Mlx-lm   Moe   Mtp   Mtplx   Ngram   Q4   Quantized   Qwen4 exp   Reasoning   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/Ecmcfee/Swift1.5-Qwen3.8-Flash-Next-q4-MTPLX 

Swift1.5 Qwen3.8 Flash Next Q4 MTPLX Parameters and Internals

LLM NameSwift1.5 Qwen3.8 Flash Next Q4 MTPLX
Repository 🤗https://huggingface.co/Ecmcfee/Swift1.5-Qwen3.8-Flash-Next-q4-MTPLX 
Base Model(s)  ukisai/Swift1.5-Qwen3.8-Flash-Next   ukisai/Swift1.5-Qwen3.8-Flash-Next
Model Size20.9b
Required VRAM72.8 GB
Updated2026-10-19
MaintainerEcmcfee
Model Typeqwen4_exp
Model Files  4.2 GB: 1-of-19   4.0 GB: 2-of-19   3.9 GB: 3-of-19   4.0 GB: 4-of-19   4.0 GB: 5-of-19   3.9 GB: 6-of-19   4.0 GB: 7-of-19   4.0 GB: 8-of-19   3.9 GB: 9-of-19   4.0 GB: 10-of-19   4.0 GB: 11-of-19   3.9 GB: 12-of-19   4.0 GB: 13-of-19   4.0 GB: 14-of-19   3.9 GB: 15-of-19   4.0 GB: 16-of-19   4.0 GB: 17-of-19   3.9 GB: 18-of-19   1.2 GB: 19-of-19   0.9 GB   1.5 GB   32.0 GB
Quantization Typeq4
Model ArchitectureQwen4ExpForConditionalGeneration
Licenseother
Model Max Length262144
Transformers Version5.8.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace