What are the hardware requirements for Moon Beam 8B A1B Preview?

Moon Beam 8B A1B Preview requires approximately 15.8 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Moon Beam 8B A1B Preview and how large is it?

Moon Beam 8B A1B Preview is developed by tablinum, 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 Moon Beam 8B A1B Preview?

Moon Beam 8B A1B Preview is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Moon Beam 8B A1B Preview by tablinum — VRAM 15.8GB, 128K context

Name: Moon Beam 8B A1B Preview
Author: tablinum

Moon Beam 8B A1B Preview is an open-source language model by tablinum. Features: 8b LLM, VRAM: 15.8GB, Context: 128K, License: mit.

Bailing hybrid Base model:finetune:inclusiona... Base model:inclusionai/ling-3.... Conversational Custom code En Engram Ling Moe Moon-beam Preview Region:us Safetensors Sharded Tablinum Tensorflow Thinking Tool-calling

Model Card on HF 🤗: https://huggingface.co/tablinum/Moon-Beam-8B-A1B-Preview

Moon Beam 8B A1B Preview Parameters and Internals

LLM Name	Moon Beam 8B A1B Preview
Repository 🤗	https://huggingface.co/tablinum/Moon-Beam-8B-A1B-Preview
Base Model(s)	inclusionAI/Ling-3.0-tiny inclusionAI/Ling-3.0-tiny
Model Size	8b
Required VRAM	15.8 GB
Updated	2026-11-06
Maintainer	tablinum
Model Type	bailing_hybrid
Model Files	4.0 GB: 1-of-4 4.0 GB: 2-of-4 4.0 GB: 3-of-4 3.8 GB: 4-of-4
Supported Languages	en
Model Architecture	BailingMoeV3ForCausalLM
License	mit
Context Length	131072
Model Max Length	131072
Transformers Version	4.57.6
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|endoftext\|>
Vocabulary Size	157184
Torch Data Type	bfloat16

Rank the Moon Beam 8B A1B Preview 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 Moon Beam 8B A1B Preview

Best Alternatives	Context / RAM	Downloads	Likes
Edge0 8B A1B Preview	128K / 0 GB	801	40
Raptor V0.5 8B A1B JANG 6M	128K / 6.8 GB	1140	8

Note: green Score (e.g. "73.2") means that the model is better than tablinum/Moon-Beam-8B-A1B-Preview.