What are the hardware requirements for Smaug Agentic?

Smaug Agentic requires approximately 650.3 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Smaug Agentic and how large is it?

Smaug Agentic is developed by abacusai, a model with 2779.9b 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 Smaug Agentic?

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

Smaug Agentic by abacusai — VRAM 650.3GB

Name: Smaug Agentic
Author: abacusai

Smaug Agentic is an open-source language model by abacusai. Features: 2779.9b LLM, VRAM: 650.3GB, License: other, LLM Explorer Score: 0.35.

Arxiv:2402.13228 8-bit Abacusai Agentic Base model:finetune:moonshotai... Base model:moonshotai/kimi-k3 Compressed-tensors Conversational Custom code Feature-extraction Image-text-to-text Kimi k3 Region:us Safetensors Sharded Smaug Tensorflow

Model Card on HF 🤗: https://huggingface.co/abacusai/Smaug-Agentic

Smaug Agentic Parameters and Internals

LLM Name	Smaug Agentic
Repository 🤗	https://huggingface.co/abacusai/Smaug-Agentic
Base Model(s)	Kimi K3 moonshotai/Kimi-K3
Model Size	2779.9b
Required VRAM	650.3 GB
Updated	2026-08-13
Maintainer	abacusai
Model Type	kimi_k3
Model Files	5.6 GB: 1-of-94 2.3 GB: 2-of-94 17.0 GB: 3-of-94 17.0 GB: 4-of-94 16.6 GB: 5-of-94 17.0 GB: 6-of-94 17.0 GB: 7-of-94 17.0 GB: 8-of-94 16.6 GB: 9-of-94 17.0 GB: 10-of-94 17.0 GB: 11-of-94 17.0 GB: 12-of-94 16.6 GB: 13-of-94 17.0 GB: 14-of-94 17.0 GB: 15-of-94 17.0 GB: 16-of-94 16.6 GB: 17-of-94 17.0 GB: 18-of-94 17.0 GB: 19-of-94 17.0 GB: 20-of-94 16.6 GB: 21-of-94 17.0 GB: 22-of-94 17.0 GB: 23-of-94 17.0 GB: 24-of-94 16.6 GB: 25-of-94 17.0 GB: 26-of-94 17.0 GB: 27-of-94 17.0 GB: 28-of-94 16.6 GB: 29-of-94 17.0 GB: 30-of-94 17.0 GB: 31-of-94 17.0 GB: 32-of-94 16.6 GB: 33-of-94 17.0 GB: 34-of-94 17.0 GB: 35-of-94 17.0 GB: 36-of-94 16.6 GB: 37-of-94 17.0 GB: 38-of-94 17.0 GB: 39-of-94 17.0 GB: 40-of-94
Model Architecture	KimiK3ForConditionalGeneration
License	other

Rank the Smaug Agentic 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 Smaug Agentic

Best Alternatives	Context / RAM	Downloads	Likes
Kimi K3	0K / 627.7 GB	493481	9307
Kimi K3	0K / 627.7 GB	18	2
Kimi K3 Abliterated V1	0K / 661.3 GB	53	3
Kimi K3 Abliterated V1	0K / 661.3 GB	25	1

Note: green Score (e.g. "73.2") means that the model is better than abacusai/Smaug-Agentic.