What are the hardware requirements for Marco Mini Instruct?

Marco Mini Instruct requires approximately 34.7 GB of VRAM and supports a context window of 32K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Marco Mini Instruct and how large is it?

Marco Mini Instruct is developed by ATH-MaaS, a model with 17.3b 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 Marco Mini Instruct?

Marco Mini Instruct is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Marco Mini Instruct by ATH-MaaS — VRAM 34.7GB, 32K context

Name: Marco Mini Instruct
Author: ATH-MaaS

Marco Mini Instruct is an open-source language model by ATH-MaaS. Features: 17.3b LLM, VRAM: 34.7GB, Context: 32K, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.26.

Ar Az Bn Conversational Cs Dataset:allenai/dolci-instruct... Dataset:bytedtsinghua-sia/dapo... Dataset:nvidia/nemotron-cascad... Dataset:nvidia/nemotron-cascad... Dataset:nvidia/nemotron-rl-ins... Dataset:nvidia/nemotron-rl-ins... Dataset:nvidia/nemotron-rl-kno... Dataset:nvidia/nemotron-rl-rea... Dataset:nvidia/nemotron-sft-mu... Dataset:skywork/skywork-or1-rl... De El En Endpoints compatible Es Fr He Hu Id Instruct It Ja Kk Ko Mixture-of-experts Moe Ms Multilingual Ne Nl On-policy distillation Pl Pt Qwen3 moe Region:us Ro Ru Safetensors Sharded Tensorflow Th Tr Uk Upcycling Ur Vi Zh

Model Card on HF 🤗: https://huggingface.co/ATH-MaaS/Marco-Mini-Instruct

Marco Mini Instruct Parameters and Internals

LLM Name	Marco Mini Instruct
Repository 🤗	https://huggingface.co/ATH-MaaS/Marco-Mini-Instruct
Model Size	17.3b
Required VRAM	34.7 GB
Updated	2026-08-30
Maintainer	ATH-MaaS
Model Type	qwen3_moe
Instruction-Based	Yes
Model Files	5.4 GB: 0-of-7 5.4 GB: 1-of-7 5.4 GB: 2-of-7 5.4 GB: 3-of-7 5.4 GB: 4-of-7 5.4 GB: 5-of-7 2.3 GB: 6-of-7
Supported Languages	en zh ar de es fr ko ja pt tr id it nl pl ru vi th he uk ms bn cs ur kk el ro hu ne az
Model Architecture	Qwen3MoeForCausalLM
License	apache-2.0
Context Length	32768
Model Max Length	32768
Transformers Version	4.57.1
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	151936
Errors	replace

Rank the Marco Mini Instruct 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 Marco Mini Instruct

Best Alternatives	Context / RAM	Downloads	Likes
Marco Mini Instruct	32K / 34.7 GB	1304	40

Note: green Score (e.g. "73.2") means that the model is better than ATH-MaaS/Marco-Mini-Instruct.