LLM EXPLORER 57,918 MODELS INDEXED

Ring Flash Linear 2.0 by inclusionAI

By inclusionAI · 33 downloads
HF Repo ↗ Copy URL

Ring Flash Linear 2.0 is an open-source language model by inclusionAI. Features: 104.2b LLM, VRAM: 207.2GB, Context: 128K, License: mit, MoE, LLM Explorer Score: 0.21.

  Arxiv:2510.19338   Bailing moe linear Base model:finetune:inclusiona... Base model:inclusionai/ling-fl...   Conversational   Custom code   En   Moe   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/inclusionAI/Ring-flash-linear-2.0 

Ring Flash Linear 2.0 Parameters and Internals

LLM NameRing Flash Linear 2.0
Repository 🤗https://huggingface.co/inclusionAI/Ring-flash-linear-2.0 
Base Model(s)  inclusionAI/Ling-flash-base-2.0   inclusionAI/Ling-flash-base-2.0
Model Size104.2b
Required VRAM207.2 GB
Updated2026-08-02
MaintainerinclusionAI
Model Typebailing_moe_linear
Model Files  8.5 GB: 1-of-31   6.6 GB: 2-of-31   6.6 GB: 3-of-31   6.6 GB: 4-of-31   6.6 GB: 5-of-31   6.5 GB: 6-of-31   6.6 GB: 7-of-31   6.6 GB: 8-of-31   6.6 GB: 9-of-31   6.6 GB: 10-of-31   6.6 GB: 11-of-31   6.6 GB: 12-of-31   6.6 GB: 13-of-31   6.5 GB: 14-of-31   6.6 GB: 15-of-31   6.6 GB: 16-of-31   6.6 GB: 17-of-31   6.6 GB: 18-of-31   6.6 GB: 19-of-31   6.6 GB: 20-of-31   6.6 GB: 21-of-31   6.5 GB: 22-of-31   6.6 GB: 23-of-31   6.6 GB: 24-of-31   6.6 GB: 25-of-31   6.6 GB: 26-of-31   6.6 GB: 27-of-31   6.6 GB: 28-of-31   6.6 GB: 29-of-31   6.5 GB: 30-of-31   7.7 GB: 31-of-31
Supported Languagesen
Model ArchitectureBailingMoeLinearV2ForCausalLM
Licensemit
Context Length131072
Model Max Length131072
Transformers Version4.56.1
Tokenizer ClassPreTrainedTokenizerFast
Padding Token<|endoftext|>
Vocabulary Size157184
Torch Data Typebfloat16

Best Alternatives to Ring Flash Linear 2.0

Best Alternatives
Context / RAM
Downloads
Likes
Ring Flash Linear 2.0 128K128K / 207.2 GB921