What are the hardware requirements for Ling 3.0 Flash Base 30T?

Ling 3.0 Flash Base 30T requires approximately 255.9 GB of VRAM and supports a context window of 8K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Ling 3.0 Flash Base 30T and how large is it?

Ling 3.0 Flash Base 30T is developed by inclusionAI, a model with 127.5b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Ling 3.0 Flash Base 30T?

Ling 3.0 Flash Base 30T is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Ling 3.0 Flash Base 30T by inclusionAI — VRAM 255.9GB, 8K context

Name: Ling 3.0 Flash Base 30T
Author: inclusionAI

Ling 3.0 Flash Base 30T is an open-source language model by inclusionAI. Features: 127.5b LLM, VRAM: 255.9GB, Context: 8K, License: mit, LLM Explorer Score: 0.28.

Arxiv:2507.17634 Bailing hybrid Custom code Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/inclusionAI/Ling-3.0-flash-base-30T

Ling 3.0 Flash Base 30T Parameters and Internals

LLM Name	Ling 3.0 Flash Base 30T
Repository 🤗	https://huggingface.co/inclusionAI/Ling-3.0-flash-base-30T
Model Size	127.5b
Required VRAM	255.9 GB
Updated	2026-08-21
Maintainer	inclusionAI
Model Type	bailing_hybrid
Model Files	10.1 GB: 1-of-27 10.1 GB: 2-of-27 10.1 GB: 3-of-27 10.1 GB: 4-of-27 8.1 GB: 5-of-27 10.1 GB: 6-of-27 10.1 GB: 7-of-27 10.1 GB: 8-of-27 10.1 GB: 9-of-27 10.1 GB: 10-of-27 8.1 GB: 11-of-27 10.1 GB: 12-of-27 8.1 GB: 13-of-27 10.1 GB: 14-of-27 8.1 GB: 15-of-27 10.1 GB: 16-of-27 10.1 GB: 17-of-27 10.1 GB: 18-of-27 10.1 GB: 19-of-27 10.1 GB: 20-of-27 10.1 GB: 21-of-27 8.1 GB: 22-of-27 10.1 GB: 23-of-27 8.9 GB: 24-of-27 10.1 GB: 25-of-27 10.7 GB: 26-of-27 3.9 GB: 27-of-27
Model Architecture	BailingMoeV3ForCausalLM
License	mit
Context Length	8192
Model Max Length	8192
Transformers Version	4.56.2
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|endoftext\|>
Vocabulary Size	157184
Torch Data Type	bfloat16

Rank the Ling 3.0 Flash Base 30T capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Ling 3.0 Flash Base 30T

Best Alternatives	Context / RAM	Downloads	Likes
Ling 3.0 Flash	256K / 254.5 GB	4189	213
Ling 3.0 Flash Int4	256K / 77.6 GB	1969	9
Ling 3.0 Flash Fin	256K / 186 GB	342	59
Ling 3.0 Flash Fp8	256K / 129.3 GB	1395	26
Ling 3.0 Flash Base Midtrain	256K / 255.9 GB	1155	5
Ling 3.0 Flash Base	256K / 255.9 GB	1192	4
Ling 3.0 Flash CIRU IU4	256K / 77.6 GB	454	5
Ling 3.0 Flash Heretic	256K / 255 GB	207	1
Ling 3.0 Flash Fin Int4	256K / 58.6 GB	47	1
Ling 3.0 Flash Fin Fp8	256K / 94.9 GB	38	1

Note: green Score (e.g. "73.2") means that the model is better than inclusionAI/Ling-3.0-flash-base-30T.