v2.2 · 2026-06-12

SmolLM2 1.7B Instruct

smollm2-1-7b-instruct

TEXT-CHAT

SmolLM2 1.7B Instruct is a 1.7B text chat model profile for local AI planning. This page records provider, license, context, quantization, hardware-fit estimates, setup hints, source links, and caveats so readers do not choose a model by name alone.

PROVIDER

HuggingFaceTB

FAMILY

SmolLM2

MODEL TYPE

text-chat

PARAMETERS

1.7B

MODALITIES

text

ARCHITECTURE

Decoder-only transformer instruction model in the SmolLM2 family.

CONTEXT WINDOW

8,192 tokens

TRAINING TOKENS

11.0T

RELEASE DATE

—

LICENSE & LINKS

Apache-2.0license text

MODEL CARD

huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct

HUGGING FACE

HuggingFaceTB/SmolLM2-1.7B-Instruct

GGUF REPOSITORIES

bartowski/SmolLM2-1.7B-Instruct-GGUF community conversion

Community GGUF conversion with Q4_K_M, Q5_K_M and Q8_0 file listings.

SETUP HINTS

OLLAMA

Run `ollama run smollm2:1.7b-instruct-q4_K_M`.

LM STUDIO

Search for a SmolLM2 1.7B Instruct GGUF such as the bartowski conversion and select Q4_K_M or Q5_K_M.

LLAMA.CPP

Use a standard text GGUF loader, for example `llama-server -hf bartowski/SmolLM2-1.7B-Instruct-GGUF:Q4_K_M`.

QUANTIZATION FILE SIZES (GGUF)

Q4_K_M

1.06 GB

Q5_K_M

1.23 GB

Q8_0

1.82 GB

Quantization note: Q4_K_M for almost all local use; Q5_K_M if you want a small quality bump with minimal extra memory.

RAM / VRAM ESTIMATES

MIN RAM

3 GB

COMFORTABLE RAM

6 GB

MIN VRAM

2 GB

COMFORTABLE VRAM

4 GB

These are conservative local-inference estimates, not official hardware requirements.

HARDWARE FIT

CPU only (no GPU)

yes

8 GB RAM

yes

16 GB RAM

yes

32 GB RAM

yes

8 GB VRAM

yes

12 GB VRAM

yes

24 GB VRAM

yes

Apple Silicon (unified memory)

yes

Hardware fit values are conservative local-inference estimates based on GGUF size plus runtime and KV-cache overhead. Actual requirements depend on context length, quantization, runtime, GPU offload, and other running apps.

BEST FOR

·Real lightweight chat within the limits of a 1.7B model

·Classification and extraction templates

·Short summarization

·Text rewriting

·Function-calling and local agent demos

·Edge-device demos

AVOID IF

·You need robust factual reliability

·You need complex legal, coding or mathematical reasoning

·You need long-document synthesis

·You need a high-quality general assistant

CAVEATS

·The exact model release date was not confirmed in cited sources from the permitted sources, so `releaseDate` is null.

·The model is primarily English and can produce inaccurate or biased outputs.

·Although the official model card describes function-calling training examples, this should be treated as lightweight tool-use suitability rather than enterprise-grade tool reliability.

·RAM and VRAM estimates are derived from GGUF/Ollama package size plus conservative overhead; KV cache, context length, batch size, runtime and GPU offload choices can materially change memory needs.

SOURCE URLS