What are the hardware requirements for G2 Nano Base?

G2 Nano Base requires approximately 0.2 GB of VRAM and supports a context window of 2K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed G2 Nano Base and how large is it?

G2 Nano Base is developed by AZERDSQ, a model with 60m 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 G2 Nano Base?

G2 Nano Base is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

G2 Nano Base by AZERDSQ — VRAM 0.2GB, 2K context

Name: G2 Nano Base
Author: AZERDSQ

G2 Nano Base is an open-source language model by AZERDSQ. Features: 60m LLM, VRAM: 0.2GB, Context: 2K, License: apache-2.0, LLM Explorer Score: 0.31.

Custom-code Custom code From-scratch G2-nano Nanogpt Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/AZERDSQ/G2-nano-base

G2 Nano Base Parameters and Internals

LLM Name	G2 Nano Base
Repository 🤗	https://huggingface.co/AZERDSQ/G2-nano-base
Model Size	60m
Required VRAM	0.2 GB
Updated	2026-09-16
Maintainer	AZERDSQ
Model Type	nanogpt
Model Files	0.2 GB
Model Architecture	NanoGPTForCausalLM
License	apache-2.0
Context Length	2048
Model Max Length	2048
Tokenizer Class	NanoGPTTokenizer
Padding Token	<pad>
Vocabulary Size	16384
Torch Data Type	float32

Rank the G2 Nano 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 G2 Nano Base

Best Alternatives	Context / RAM	Downloads	Likes
G2 Nano Instruct	2K / 0.2 GB	274	1
G1 Nano Instruct	2K / 0.2 GB	112	2

Note: green Score (e.g. "73.2") means that the model is better than AZERDSQ/G2-nano-base.