---
title: BAAI/bge-reranker-v2-m3
description: "More details please refer to our Github: FlagEmbedding.. XLM-RoBERTa, 568M parameters."
canonical_url: https://superlinked.com/models/baai-bge-reranker-v2-m3
last_updated: 2026-06-10
---

# BAAI/bge-reranker-v2-m3

More details please refer to our Github: FlagEmbedding.

Source: [BAAI/bge-reranker-v2-m3 on HuggingFace](https://huggingface.co/BAAI/bge-reranker-v2-m3)

## Overview

| Field | Value |
|-------|-------|
| Architecture | XLM-RoBERTa |
| Parameters | 568M |
| Tasks | Score |
| Outputs | Score |
| Max sequence length | 8,192 tokens |
| License | apache-2.0 |
| Inputs | text |
| Languages | multilingual |

## Benchmarks

### AskUbuntuDupQuestions

Domain: technology · Task: reranking · Language: en

Duplicate question detection from AskUbuntu

Corpus: 6,743 · Queries: 360

**Quality:** ndcg at 10: 0.6763 · map at 10: 0.5245 · mrr at 10: 0.7611

**Performance (L4 b1 c16):** Query 6.5K tok/s · Query p50 40.6ms

[Reference](https://github.com/taolei87/askubuntu)

### CMedQAv1Reranking

Domain: medical · Task: reranking · Language: zh

Chinese medical question answering reranking (v1)

Corpus: 100,000 · Queries: 2,000

**Quality:** map at 10: 0.8192 · mrr at 10: 0.8550

[Reference](https://github.com/zhangsheng93/cMedQA)

### CMedQAv2Reranking

Domain: medical · Task: reranking · Language: zh

Chinese medical question answering reranking (v2)

Corpus: 108,000 · Queries: 4,000

**Quality:** map at 10: 0.8287 · mrr at 10: 0.8608

[Reference](https://github.com/zhangsheng93/cMedQA2)

### CQADupstackPhysicsRetrieval?candidates_model=Alibaba-NLP

Domain: general · Task: retrieval · Language: en

**Performance (L4 b1 c16):** Query 26.5K tok/s · Query p50 89.8ms

### CosQA?candidates_model=Alibaba-NLP

Domain: general · Task: retrieval · Language: en

**Performance (L4 b1 c16):** Query 13.9K tok/s · Query p50 66.1ms

### FiQA2018?candidates_model=Alibaba-NLP

Domain: general · Task: retrieval · Language: en

**Performance (L4 b1 c16):** Query 30.0K tok/s · Query p50 90.2ms

### LegalBenchConsumerContractsQA?candidates_model=Alibaba-NLP

Domain: general · Task: retrieval · Language: en

**Performance (L4 b1 c16):** Query 51.3K tok/s · Query p50 93.5ms

### MMarcoReranking

Domain: general · Task: reranking · Language: zh

Multilingual MARCO passage reranking (Chinese)

**Quality:** map at 10: 0.3354 · mrr at 10: 0.3401

[Reference](https://arxiv.org/abs/2304.03679)

### NFCorpus

Domain: medical · Task: retrieval · Language: en

Biomedical literature search from NutritionFacts.org

Corpus: 3,593 · Queries: 323

**Variant: default_bm25-k-100**

**Quality:** ndcg at 10: 0.3285 · map at 10: 0.2339 · mrr at 10: 0.5301

**Variant: default_bm25-k-200_qrels-k-50**

**Quality:** ndcg at 10: 0.3674 · map at 10: 0.2602 · mrr at 10: 0.5765

**Variant: default_bm25-k-200**

**Quality:** ndcg at 10: 0.3280 · map at 10: 0.2331 · mrr at 10: 0.5264

**Variant: default_bm25-k-50_qrels-k-10**

**Quality:** ndcg at 10: 0.4029 · map at 10: 0.2832 · mrr at 10: 0.6087

**Variant: default_bm25-k-50**

**Quality:** ndcg at 10: 0.3306 · map at 10: 0.2375 · mrr at 10: 0.5380

**Variant: default_bm25-k-100_qrels-k-10**

**Quality:** ndcg at 10: 0.3759 · map at 10: 0.2621 · mrr at 10: 0.5876

**Variant: default_bm25-k-50_qrels-k-50**

**Quality:** ndcg at 10: 0.4243 · map at 10: 0.3065 · mrr at 10: 0.6220

**Variant: default_bm25-k-200_qrels-k-10**

**Quality:** ndcg at 10: 0.3557 · map at 10: 0.2497 · mrr at 10: 0.5677

**Variant: default_bm25-k-100_qrels-k-50**

**Quality:** ndcg at 10: 0.3930 · map at 10: 0.2792 · mrr at 10: 0.5996

[Reference](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/)

### NFCorpus?candidates_model=Alibaba-NLP

Domain: general · Task: retrieval · Language: en

**Performance (L4 b1 c16):** Query 35.3K tok/s · Query p50 109.6ms

### SCIDOCS?candidates_model=Alibaba-NLP

Domain: general · Task: retrieval · Language: en

**Performance (L4 b1 c16):** Query 29.4K tok/s · Query p50 98.7ms

### SciFact?candidates_model=Alibaba-NLP

Domain: general · Task: retrieval · Language: en

**Performance (L4 b1 c16):** Query 33.4K tok/s · Query p50 113.2ms

### StackOverflowQA?candidates_model=Alibaba-NLP

Domain: general · Task: retrieval · Language: en

**Performance (L4 b1 c16):** Query 53.1K tok/s · Query p50 111.1ms

### T2Reranking

Domain: general · Task: reranking · Language: zh

Chinese passage ranking benchmark

**Quality:** map at 10: 0.5630 · mrr at 10: 0.7781

[Reference](https://arxiv.org/abs/2304.03679)