LLM EXPLORER 59,182 MODELS INDEXED

LongCat Flash Lite Int4 AutoRound by Intel

By Intel · 74 downloads
HF Repo ↗ Copy URL

LongCat Flash Lite Int4 AutoRound is an open-source language model by Intel. Features: 37.9b LLM, VRAM: 83.7GB, Context: 320K, LLM Explorer Score: 0.23.

  Arxiv:2309.05516   4-bit   Auto-round Base model:meituan-longcat/lon... Base model:quantized:meituan-l...   Conversational   Custom code   Longcat flash ngram   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/Intel/LongCat-Flash-Lite-int4-AutoRound 

LongCat Flash Lite Int4 AutoRound Parameters and Internals

LLM NameLongCat Flash Lite Int4 AutoRound
Repository 🤗https://huggingface.co/Intel/LongCat-Flash-Lite-int4-AutoRound 
Base Model(s)  LongCat Flash Lite   meituan-longcat/LongCat-Flash-Lite
Model Size37.9b
Required VRAM83.7 GB
Updated2026-07-31
MaintainerIntel
Model Typelongcat_flash_ngram
Model Files  4.3 GB: 1-of-19   4.3 GB: 2-of-19   4.3 GB: 3-of-19   4.3 GB: 4-of-19   2.5 GB: 5-of-19   0.8 GB: 6-of-19   5.2 GB: 7-of-19   5.2 GB: 8-of-19   5.2 GB: 9-of-19   5.2 GB: 10-of-19   5.2 GB: 11-of-19   5.2 GB: 12-of-19   5.2 GB: 13-of-19   5.2 GB: 14-of-19   5.2 GB: 15-of-19   5.2 GB: 16-of-19   5.2 GB: 17-of-19   5.2 GB: 18-of-19   0.8 GB: 19-of-19
Model ArchitectureLongcatFlashNgramForCausalLM
Context Length327680
Model Max Length327680
Transformers Version4.57.6
Tokenizer ClassBloomTokenizer
Padding Token<longcat_pad>
Vocabulary Size131072