What are the hardware requirements for LLaDA 8B Base?

LLaDA 8B Base requires approximately 16.1 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed LLaDA 8B Base and how large is it?

LLaDA 8B Base is developed by GSAI-ML, a model with 8b 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 LLaDA 8B Base?

LLaDA 8B Base is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

LLaDA 8B Base by GSAI-ML — VRAM 16.1GB

Name: LLaDA 8B Base
Author: GSAI-ML

LLaDA 8B Base is an open-source language model by GSAI-ML. Features: 8b LLM, VRAM: 16.1GB, License: mit, LLM Explorer Score: 0.24.

Conversational Custom code Llada Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/GSAI-ML/LLaDA-8B-Base

LLaDA 8B Base Parameters and Internals

LLM Name	LLaDA 8B Base
Repository 🤗	https://huggingface.co/GSAI-ML/LLaDA-8B-Base
Model Size	8b
Required VRAM	16.1 GB
Updated	2026-08-01
Maintainer	GSAI-ML
Model Type	llada
Model Files	2.1 GB: 1-of-6 3.0 GB: 2-of-6 3.0 GB: 3-of-6 3.0 GB: 4-of-6 2.9 GB: 5-of-6 2.1 GB: 6-of-6
Model Architecture	LLaDAModelLM
License	mit
Transformers Version	4.46.3
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|endoftext\|>
Vocabulary Size	126464

Rank the LLaDA 8B Base 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 LLaDA 8B Base

Best Alternatives	Context / RAM	Downloads	Likes
LLaDA 8B Instruct	0K / 16.1 GB	419335	359
LLaDA 1.5	0K / 16.1 GB	75107	41

Note: green Score (e.g. "73.2") means that the model is better than GSAI-ML/LLaDA-8B-Base.