What are the hardware requirements for Phi 3 Mini 4K Instruct?

Phi 3 Mini 4K Instruct requires approximately 7.6 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 Phi 3 Mini 4K Instruct and how large is it?

Phi 3 Mini 4K Instruct is developed by unsloth, a model with 3.8b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Phi 3 Mini 4K Instruct perform on standard benchmarks?

Phi 3 Mini 4K Instruct has the following published scores: MMLU 69.08, LMSYS ELO 1128. Compare against reference models on this page or on the LLM Explorer leaderboards.

Phi 3 Mini 4K Instruct by unsloth — VRAM 7.6GB, 4K context

Name: Phi 3 Mini 4K Instruct
Author: unsloth

Phi 3 Mini 4K Instruct is an open-source language model by unsloth. Features: 3.8b LLM, VRAM: 7.6GB, Context: 4K, License: mit, Instruction-Based, LLM Explorer Score: 0.34, ELO: 1128, Arc: 63, HellaSwag: 80.6, MMLU: 69.1, GSM8K: 73.9.

Conversational En Endpoints compatible Instruct Mistral Phi Phi3 Region:us Safetensors Sharded Tensorflow Unsloth

Model Card on HF 🤗: https://huggingface.co/unsloth/Phi-3-mini-4k-instruct

Phi 3 Mini 4K Instruct Benchmarks

LMSys ELO 1128 vs 1272 (so35) -11.3%

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

Phi 3 Mini 4K Instruct Parameters and Internals

Model Type

text generation, transformers

Use Cases

Areas:

research, commercial applications

Primary Use Cases:

text generation, model finetuning

Additional Notes

The model can be directly quantized to 4bit using 'bitsandbytes'.

LLM Name	Phi 3 Mini 4K Instruct
Repository 🤗	https://huggingface.co/unsloth/Phi-3-mini-4k-instruct
Model Size	3.8b
Required VRAM	7.6 GB
Updated	2026-07-22
Maintainer	unsloth
Model Type	mistral
Instruction-Based	Yes
Model Files	5.0 GB: 1-of-2 2.6 GB: 2-of-2
Supported Languages	en
Model Architecture	MistralForCausalLM
License	mit
Context Length	4096
Model Max Length	4096
Transformers Version	4.44.2
Tokenizer Class	LlamaTokenizer
Padding Token	<\|placeholder6\|>
Vocabulary Size	32064
Torch Data Type	bfloat16

Rank the Phi 3 Mini 4K Instruct 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

Quantized Models of the Phi 3 Mini 4K Instruct

Model	Downloads	VRAM
Scm Phi3 Q4 K M V3	19	2 GB
Scm Phi3 Q8 V3	14	4 GB
Scm Phi3 Q5 K M V3	10	2 GB
Scm Phi3 Q8	6	4 GB
Scm Phi3 Q4 K M	5	2 GB
Scm Phi3 16 V3	5	7 GB
...in 2.9.1 Phi 3 Kensho 4.5B AWQ	7	3 GB
...Sentiment May 24 2024 2epoches	20	7 GB
Dj Phi 3 F16 GGUF	21	7 GB
Dj Phi 3 3.8B 16bit	18	7 GB

Best Alternatives to Phi 3 Mini 4K Instruct

Best Alternatives	Context / RAM	Downloads	Likes
Llama Phi 3 DoRA	4K / 7.6 GB	26	6
Empty Phi 3 Mini	4K / 7.6 GB	10	0
...ling Assistant Spanish Pofi V2	4K / 7.7 GB	15	0
...i3 Short Story Merged Bfloat16	4K / 7.6 GB	10	1
Phi 3 Mini 4K Instruct Orpo V2	4K / 14.9 GB	5	1
Tonmai GPT	4K / 7.6 GB	314	0
Wandaa V02 Merged	4K / 7.6 GB	28	0
... Instruct Unsloth Merged 16bit	4K / 7.6 GB	14	2
...ent V56 3 W Unknown 5 Lr 0.002	4K / 7.6 GB	7	0
...Word Problems 200K Model 16bit	4K / 7.6 GB	3879	0