HuggingFaceTB / text-generation updated 1 year ago

SmolLM3-3B

Name: SmolLM3-3B
Author: HuggingFaceTB

1. Model Summary 2. How to use 3. Evaluation 4. Training 5. Limitations 6. License

Params

3.1 B

Context

65,536

Downloads 30d

614 K

Likes

1,034

Commercial use: allowed apache-2.0 Not gated SAFETENSORS 8 languages View on Hugging Face ↗

Download history

daily snapshots · 56 days

▲ 169 K in the last 30 days (21.6%)

872 K558 K

Jul 28Aug 15Sep 3Sep 21

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	bf16	6.2 GB	7.7 GB	✅ Runs comfortably
model.safetensors (bf16, full)	bf16 + 64K ctx	6.2 GB	11.0 GB	✅ Runs comfortably

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Run it

copy-paste, exact tags checked against the Hub

~ · curl · api/v1

$ curl -s https://aimodelscomparison.com/api/v1/models/smollm3-3b
{
  "hf_id": "HuggingFaceTB/SmolLM3-3B",
  "params_b": 3.08,
  "context_length": 65536,
  "license": { "id": "apache-2.0", "commercial": "yes" },
  "downloads_30d": 613886,
  "vram_estimates": [
    { "quant": "bf16", "gb": 7.7 }
  ],
  "updated_at": "2026-07-28T18:04:04Z"
}

est. VRAM —on RTX 4090 · 24 GBJSON API →

Specifications

Architecture: SmolLM3ForCausalLM
Parameters: 3.1 B
Tensor type: BF16
Context length: 65,536
Vocabulary: 128,256
Layers / heads: 36 / 16
Licence: apache-2.0
First seen on the Hub: 2025-07-08
Training datasets: undisclosed
Added to our catalog: 2026-07-28

Family

Base model and the most-downloaded derivatives in the catalog.

Derivative TwIL-LM3 webAI-Official 384 K

Compare with any text-generation model

Popular comparisons

Qwen3-8B 12.9 M Qwen2.5-7B-Instruct 9.7 M Qwen2.5-1.5B-Instruct 7.4 M Qwen3-4B 6.9 M Qwen2.5-3B-Instruct 4.8 M Qwen3-4B-Instruct-2507 4.0 M