HuggingFaceTB/SmolLM3-3B
Parameters3.1B
Context64K
Input → outputtext → text
Licenseapache-2.0
PrecisionBF16
ArchitectureSmolLM3ForCausalLM
ReleasedJul 8, 2025
Downloads711.7K
Based on HuggingFaceTB/SmolLM3-3B-Base
Other sizes and formats in the family
| Model | Params | Precision | Downloads |
|---|---|---|---|
| HuggingFaceTB/SmolLM3-3B-Base | 3.1B | BF16 | 97.8K |
| HuggingFaceTB/SmolLM3-3B-ONNXquantized | 3B | – | 474 |
Built on this model
| Model | Params | Precision | Downloads |
|---|---|---|---|
| HuggingFaceTB/SmolLM3-3B-ONNXquantized | 3B | – | 474 |
This release in the news
No coverage of this exact release yet.
Recent news about SmolLM
- Oct 11, 2026 · Open-source releaseConverting dense models into Mixture-of-Experts
- Oct 7, 2026 · Research paperSemanticFold: Latent Sequence Compression SeparatesLanguage Modeling, Decodability, and Reasoning
- Oct 7, 2026 · Research paperBoT-GRPO: Efficient Process-Reward RL for Reasoning via Bag-of-Token Aggregation
- Oct 7, 2026 · Research paperEvaluating Trajectory Features for Routing Final-Layer Attention