What are the hardware requirements for ReasoningCore 3B Instruct R01 Reflect?

ReasoningCore 3B Instruct R01 Reflect requires approximately 6.5 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 ReasoningCore 3B Instruct R01 Reflect and how large is it?

ReasoningCore 3B Instruct R01 Reflect is developed by EpistemeAI, a model with 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 ReasoningCore 3B Instruct R01 Reflect?

ReasoningCore 3B Instruct R01 Reflect is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

ReasoningCore 3B Instruct R01 Reflect by EpistemeAI — VRAM 6.5GB, 128K context

Name: ReasoningCore 3B Instruct R01 Reflect
Author: EpistemeAI

ReasoningCore 3B Instruct R01 Reflect is an open-source language model by EpistemeAI. Features: 3b LLM, VRAM: 6.5GB, Context: 128K, License: proprietary, Quantized, Instruction-Based, LLM Explorer Score: 0.23.

4bit Autotrain compatible Conversational En Endpoints compatible Instruct Llama Quantized Region:us Safetensors Sharded Tensorflow Trl Unsloth

Model Card on HF 🤗: https://huggingface.co/EpistemeAI/ReasoningCore-3B-Instruct-r01-Reflect

ReasoningCore 3B Instruct R01 Reflect Parameters and Internals

LLM Name	ReasoningCore 3B Instruct R01 Reflect
Repository 🤗	https://huggingface.co/EpistemeAI/ReasoningCore-3B-Instruct-r01-Reflect
Base Model(s)	unsloth/llama-3.2-3b-instruct-unsloth-bnb-4bit unsloth/llama-3.2-3b-instruct-unsloth-bnb-4bit
Model Size	3b
Required VRAM	6.5 GB
Updated	2025-05-30
Maintainer	EpistemeAI
Model Type	llama
Instruction-Based	Yes
Model Files	5.0 GB: 1-of-2 1.5 GB: 2-of-2
Supported Languages	en
Gated Model	Yes
Quantization Type	4bit
Model Architecture	LlamaForCausalLM
License	proprietary
Context Length	131072
Model Max Length	131072
Transformers Version	4.48.3
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	<\|finetune_right_pad_id\|>
Vocabulary Size	128256
Torch Data Type	bfloat16

Rank the ReasoningCore 3B Instruct R01 Reflect 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 ReasoningCore 3B Instruct R01 Reflect

Best Alternatives	Context / RAM	Downloads	Likes
Llama3 Finetuned Id 16bit	128K / 6.5 GB	652	0
Slm Rag Sft Merged	128K / 6.5 GB	297	0
Llama3.2 3B Legal Id Ft	128K / 6.5 GB	264	0
...lama 3.2 3B Legal Assistant Id	128K / 6.5 GB	283	0
Fine Tuning LLM	128K / 6.5 GB	534	0
...Llama 3.2 3B Legal Chatbot Sft	128K / 6.5 GB	817	0
Cheese 3B	128K / 1.8 GB	249	1
Legal Llama3.2 3B Sft Id	128K / 6.5 GB	680	0
...gabl Legal Chatbot Llama3.2 3B	128K / 6.5 GB	528	0
Llama 3.2 3B Legal Id Sft	128K / 6.5 GB	236	0

Note: green Score (e.g. "73.2") means that the model is better than EpistemeAI/ReasoningCore-3B-Instruct-r01-Reflect.