BGE Reranker v2 M3
Model ID: vllm/release/bge-reranker-v2-m3 BAAI Experimental
EU AI Act
Section titled “EU AI Act”| Attribute | Value |
|---|---|
| Classification | Not a GPAI model (specialised model) |
| Risk class | Limited Risk (Article 50) |
| Transparency obligation | Yes |
| noris role | Distributor |
As a specialised single-purpose model for embeddings or reranking, this model does not fall under the definition of a GPAI model (Article 3(63)). As a component of your AI system, it is subject to the general transparency obligations (Article 50). noris acts as a distributor.
Technical Specifications
Section titled “Technical Specifications”| Property | Value |
|---|---|
| Model name | BGE Reranker v2 M3 |
| Provider | BAAI |
| Model ID | vllm/release/bge-reranker-v2-m3 |
| Parameters | 0.6B |
| Architecture | Cross-encoder (XLM-RoBERTa) |
| Languages | Multilingual |
| Max input length | 512 tokens |
| License | Apache 2.0 |
Recommended Use Cases
Section titled “Recommended Use Cases”- Reranking in RAG pipelines
- Relevance sorting
- Document ranking
Known Limitations
Section titled “Known Limitations”Risk Assessment (EU AI Act)
Section titled “Risk Assessment (EU AI Act)”| Risk Factor | Category | Level |
|---|---|---|
| Model type | No generative output | Low |
| Hallucinations | Not applicable | Low |
| Overall | Small model, specialized task | Low |
Original Sources
Section titled “Original Sources”- HuggingFace: https://huggingface.co/BAAI/bge-reranker-v2-m3
- License: Apache 2.0
Model card created on July 2, 2026. This model card serves EU AI Act transparency purposes (Regulation (EU) 2024/1689). The information is based on publicly available data from the model provider.
