What are the hardware requirements for MAXWELL?

MAXWELL requires approximately 3.1 GB of VRAM and supports a context window of 4K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed MAXWELL and how large is it?

MAXWELL is developed by Xerv-AI, a model with 1.5b 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 MAXWELL?

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

MAXWELL by Xerv-AI — VRAM 3.1GB, 4K context

Name: MAXWELL
Author: Xerv-AI

MAXWELL is an open-source language model by Xerv-AI. Features: 1.5b LLM, VRAM: 3.1GB, Context: 4K, License: apache-2.0, Quantized, Instruction-Based, Merged, LLM Explorer Score: 0.23.

Merged Model 4bit Analytical Base model:finetune:unsloth/qw... Base model:unsloth/qwen2.5-mat... Conversational Dataset:xerv-ai/tart En Endpoints compatible Eval-results Fp16 Instruct Logic Mathematics Meta-aggregation Model-index Physics Quantized Qwen2 Qwen2.5-math Reasoning Region:us Safetensors Science Stem Stss-framework Unsloth

Model Card on HF 🤗: https://huggingface.co/Xerv-AI/MAXWELL

MAXWELL Parameters and Internals

LLM Name	MAXWELL
Repository 🤗	https://huggingface.co/Xerv-AI/MAXWELL
Model Name	MAXWELL
Model Creator	Xerv-AI
Base Model(s)	unsloth/Qwen2.5-Math-1.5B-Instruct-bnb-4bit unsloth/Qwen2.5-Math-1.5B-Instruct-bnb-4bit
Merged Model	Yes
Model Size	1.5b
Required VRAM	3.1 GB
Updated	2026-06-30
Maintainer	Xerv-AI
Model Type	qwen2
Instruction-Based	Yes
Model Files	3.1 GB
Supported Languages	en
Quantization Type	fp16\|4bit
Model Architecture	Qwen2ForCausalLM
License	apache-2.0
Context Length	4096
Model Max Length	4096
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|PAD_TOKEN\|>
Vocabulary Size	151936
Torch Data Type	float16
Errors	replace

Rank the MAXWELL 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 MAXWELL

Best Alternatives	Context / RAM	Downloads	Likes
Quantrexa Prime 1.5B	32K / 3.1 GB	497	1
Qwen25 1.5B Alpaca Id Sft	32K / 3.1 GB	2038	0
....5 1.5B Instruct Alpaca Id Sft	32K / 3.1 GB	332	0
Qwen2.5 1.5B Instruct Legal Id	32K / 3.1 GB	1425	0
Vera Qwen 1.5B DPO	32K / 3.1 GB	381	1
Legal Chatbot Finetuned	32K / 3.1 GB	341	0
Legal Assistant Sft	32K / 3.1 GB	312	1
Legal Chatbot Llama3 Finetuned	32K / 3.1 GB	389	0
NOVA Tendo 3.8	32K / 3.1 GB	19	1
Legal Chatbot Llama3 Finetuned	32K / 3.1 GB	297	0

Note: green Score (e.g. "73.2") means that the model is better than Xerv-AI/MAXWELL.