What are the hardware requirements for Laguna S 2.1 OQ5e?

Laguna S 2.1 OQ5e requires approximately 78.3 GB of VRAM and supports a context window of 1024K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Laguna S 2.1 OQ5e and how large is it?

Laguna S 2.1 OQ5e is developed by mlx-community, a model with 20.5b 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 Laguna S 2.1 OQ5e?

Laguna S 2.1 OQ5e is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Laguna S 2.1 OQ5e by mlx-community — VRAM 78.3GB, 1024K context

Name: Laguna S 2.1 OQ5e
Author: mlx-community

Laguna S 2.1 OQ5e is an open-source language model by mlx-community. Features: 20.5b LLM, VRAM: 78.3GB, Context: 1024K, License: openmdw-1.1, LLM Explorer Score: 0.3.

5-bit Base model:poolside/laguna-s-2... Base model:quantized:poolside/... Conversational Custom code Laguna Mlx Moe Oq Quantized Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/mlx-community/Laguna-S-2.1-oQ5e

Laguna S 2.1 OQ5e Parameters and Internals

LLM Name	Laguna S 2.1 OQ5e
Repository 🤗	https://huggingface.co/mlx-community/Laguna-S-2.1-oQ5e
Base Model(s)	poolside/Laguna-S-2.1 poolside/Laguna-S-2.1
Model Size	20.5b
Required VRAM	78.3 GB
Updated	2026-08-02
Maintainer	mlx-community
Model Type	laguna
Model Files	5.2 GB: 1-of-15 5.3 GB: 2-of-15 5.3 GB: 3-of-15 5.3 GB: 4-of-15 5.3 GB: 5-of-15 5.3 GB: 6-of-15 5.3 GB: 7-of-15 5.3 GB: 8-of-15 5.3 GB: 9-of-15 5.3 GB: 10-of-15 5.3 GB: 11-of-15 5.3 GB: 12-of-15 5.3 GB: 13-of-15 5.3 GB: 14-of-15 4.2 GB: 15-of-15
Model Architecture	LagunaForCausalLM
License	openmdw-1.1
Context Length	1048576
Model Max Length	1048576
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	〈\|PAD\|〉
Vocabulary Size	100352
Torch Data Type	bfloat16

Rank the Laguna S 2.1 OQ5e 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