What are the hardware requirements for EVA Instruct 32B EXL2 4.0?

EVA Instruct 32B EXL2 4.0 requires approximately 17.7 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed EVA Instruct 32B EXL2 4.0 and how large is it?

EVA Instruct 32B EXL2 4.0 is developed by ParasiticRogue, a model with 32b 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 EVA Instruct 32B EXL2 4.0?

EVA Instruct 32B EXL2 4.0 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

EVA Instruct 32B EXL2 4.0 by ParasiticRogue — VRAM 17.7GB, 128K context

Name: EVA Instruct 32B EXL2 4.0
Author: ParasiticRogue

EVA Instruct 32B EXL2 4.0 is an open-source language model by ParasiticRogue. Features: 32b LLM, VRAM: 17.7GB, Context: 128K, License: apache-2.0, Quantized, Instruction-Based, Merged, LLM Explorer Score: 0.15.

Merged Model 4-bit Autotrain compatible Base model:eva-unit-01/eva-qwe... Base model:qwen/qwen2.5-32b-in... Chat Conversational En Endpoints compatible Exl2 Instruct Not-for-all-audiences Quantized Qwen2 Region:us Roleplay Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/ParasiticRogue/EVA-Instruct-32B-exl2-4.0

EVA Instruct 32B EXL2 4.0 Parameters and Internals

LLM Name	EVA Instruct 32B EXL2 4.0
Repository 🤗	https://huggingface.co/ParasiticRogue/EVA-Instruct-32B-exl2-4.0
Base Model(s)	Qwen/Qwen2.5-32B-Instruct EVA-UNIT-01/EVA-Qwen2.5-32B-v0.0 Qwen/Qwen2.5-32B-Instruct EVA-UNIT-01/EVA-Qwen2.5-32B-v0.0
Merged Model	Yes
Model Size	32b
Required VRAM	17.7 GB
Updated	2025-09-23
Maintainer	ParasiticRogue
Model Type	qwen2
Instruction-Based	Yes
Model Files	4.2 GB: 1-of-5 4.3 GB: 2-of-5 4.2 GB: 3-of-5 4.2 GB: 4-of-5 0.8 GB: 5-of-5
Supported Languages	en
Quantization Type	exl2
Model Architecture	Qwen2ForCausalLM
License	apache-2.0
Context Length	131072
Model Max Length	131072
Transformers Version	4.44.0
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	152064
Torch Data Type	bfloat16
Errors	replace

Rank the EVA Instruct 32B EXL2 4.0 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 EVA Instruct 32B EXL2 4.0

Best Alternatives	Context / RAM	Downloads	Likes
....5 Coder 32B Instruct Bnb 4bit	32K / 19.2 GB	5519	6
...wen2.5 Coder 32B Instruct 3bit	32K / 14.3 GB	144	4
Qwen 32B Incorrect Trivia 2	32K / 65.8 GB	12	0
Qwen2.5 32B Instruct Bnb 4bit	32K / 19.2 GB	49427	15
...wen2.5 Coder 32B Instruct 4bit	32K / 18.5 GB	2357	12
Cogito R1	32K / 65.8 GB	8	5
...wen2.5 Coder 32B Instruct 8bit	32K / 34.9 GB	934	14
Qwen2.5 32B Instruct 4bit	32K / 18.5 GB	1502	4
Qwen2.5 32B Instruct 8bit	32K / 34.9 GB	979	7
...truct Gptqmodel 4bit Vortex V1	32K / 21.3 GB	83	16

Note: green Score (e.g. "73.2") means that the model is better than ParasiticRogue/EVA-Instruct-32B-exl2-4.0.