SmolLM3 — local AI model by Hugging Face
Written by Jakub Rusinowski · Last updated
Hugging Face's fully-open 3B: weights, training data, recipe and 100+ intermediate checkpoints all published. Dual-mode — a deep-thinking path for hard problems and a fast path for everything else — with 128K context and six natively supported languages. The reference model for anyone fine-tuning at the small end, which makes it the natural companion to the datasets hub.
Variants
The smallest SmolLM3 variant needs about 3 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache.
| Model | VRAM at Q4 | VRAM | Context | Run it |
|---|---|---|---|---|
| SmolLM3 3B → 3B | ~2.7 GB | 131,072 | ollama run smollm3:3b |
Memory is quantized weights plus overhead at Q4_K_M, from the same engine as the GPU & VRAM checker.
How to run SmolLM3 locally
Install Ollama, then pull the tag.
ollama run smollm3:3bPick a size above for its own VRAM figure, speed estimate and install command.
Licence
Commercial use permitted. No usage restrictions beyond attribution.
Applies to: SmolLM3 3BRecommended GPU
The cheapest catalogued GPU that runs SmolLM3 locally (min 3 GB VRAM) is the Intel Arc B570 (10 GB).
Can I run SmolLM3 on my GPU?
SmolLM3 — frequently asked questions
How much VRAM does SmolLM3 need?
SmolLM3 needs about 3 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: SmolLM3 3B (3 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Can I run SmolLM3 on an RTX 4090 (24 GB)?
Yes — SmolLM3 runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
What quantization should I use for SmolLM3?
Q4_K_M is the best balance of quality and VRAM for SmolLM3 in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
How do I run SmolLM3 with Ollama?
Install Ollama, then run: ollama run smollm3:3b. This downloads SmolLM3 and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.