What are the hardware requirements for SmolLM2 135M Instruct?

SmolLM2 135M Instruct requires approximately 0.3 GB of VRAM and supports a context window of 8K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed SmolLM2 135M Instruct and how large is it?

SmolLM2 135M Instruct is developed by HuggingFaceTB, a model with 135m 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 SmolLM2 135M Instruct?

SmolLM2 135M Instruct is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

SmolLM2 135M Instruct by HuggingFaceTB — VRAM 0.3GB, 8K context

Name: SmolLM2 135M Instruct
Author: HuggingFaceTB

SmolLM2 135M Instruct is an open-source language model by HuggingFaceTB. Features: 135m LLM, VRAM: 0.3GB, Context: 8K, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.28.

Arxiv:2502.02737 Base model:huggingfacetb/smoll... Base model:quantized:huggingfa... Conversational En Endpoints compatible Instruct Llama Onnx Region:us Safetensors Tensorboard Transformers.js

Model Card on HF 🤗: https://huggingface.co/HuggingFaceTB/SmolLM2-135M-Instruct

SmolLM2 135M Instruct Parameters and Internals

Model Type

language model, instruction following

Use Cases

Areas:

assistive tools

Applications:

text rewriting, summarization, function calling

Primary Use Cases:

Assistive tools for generating English text

Limitations:

Models generate English content, Factually inaccuracy possible, Logical inconsistencies and biases

Considerations:

Verify important information and critically evaluate.

Additional Notes

SmolLM2 models offer lightweight capabilities for on-device use.

Supported Languages

en (primary)

Training Details

Data Sources:

FineWeb-Edu, DCLM, The Stack, UltraFeedback, Synth-APIGen-v0.1

Data Volume:

2T tokens

Methodology:

Supervised fine-tuning (SFT), Direct Preference Optimization (DPO)

Hardware Used:

64 H100 GPUs

Model Architecture:

Transformer decoder

Input Output

Accepted Modalities:

text

LLM Name	SmolLM2 135M Instruct
Repository 🤗	https://huggingface.co/HuggingFaceTB/SmolLM2-135M-Instruct
Base Model(s)	SmolLM2 135M HuggingFaceTB/SmolLM2-135M
Model Size	135m
Required VRAM	0.3 GB
Updated	2026-08-08
Maintainer	HuggingFaceTB
Model Type	llama
Instruction-Based	Yes
Model Files	0.3 GB 0.0 GB
Supported Languages	en
Context Length	8k
Model Architecture	LlamaForCausalLM
License	apache-2.0
Context Length	8192
Model Max Length	8192
Transformers Version	4.42.3
Tokenizer Class	GPT2Tokenizer
Padding Token	<\|im_end\|>
Vocabulary Size	49152
Torch Data Type	bfloat16

Rank the SmolLM2 135M 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 SmolLM2 135M Instruct

Model	Likes	Downloads	VRAM
...5M Instruct Bnb 4bit Reasoning	1	36	0 GB
SmolLM2 135M Instruct Q8 Mlx	2	198	0 GB

Best Alternatives to SmolLM2 135M Instruct

Best Alternatives	Context / RAM	Downloads	Likes
SmolLM2 135M Reasoning 5K	8K / 0.3 GB	276	3
... Subtitle Postprocessor Merged	8K / 0.5 GB	16	1
SmolLM2 135M Instruction Tuned	8K / 0.3 GB	10	0
SmolLM2 Rethink 135M	8K / 0.5 GB	6	2
SmolLM2 135M Instruct	8K / 0.3 GB	4887	3
SmolLM2 135M Instruct Ita	8K / 0.1 GB	24	0
...molLM2 135M Instruct Reasoning	8K / 0.3 GB	7	0
...wre324 R1 SmolLM2 135M Distill	8K / 0.5 GB	8	0
...rtis SmolLM2 135M Instruct DPO	8K / 0.5 GB	21	0
Kurtis SmolLM2 135M Instruct	8K / 0.5 GB	20	0