Compare models
Pick two or three models to compare license, hardware, and live stats side by side.
| BGE-M3 | all-MiniLM-L6-v2 | |
|---|---|---|
| Org | BAAI | Sentence-Transformers |
| Params | 568M | 22M |
| Modality | Embedding | Embedding |
| License | PermissiveMIT | PermissiveApache 2.0 |
| Hardware | Consumer GPU | Consumer GPU |
| Released | 2024-01 | 2021-08 |
| Take | A versatile embedding model that unifies dense, sparse, and multi-vector (ColBERT-style) retrieval in one checkpoint across 100+ languages. | The default lightweight embedding baseline for years: tiny, extremely fast, and good enough for most semantic search prototypes. |
| Downloads | — | — |
| Likes | — | — |