Skip to content

Bge m3

Model Overview

BGE-M3 is distinguished for its versatility in Multi-Functionality, Multi-Linguality, and Multi-Granularity.

Multi-Functionality: It can simultaneously perform the three common retrieval functionalities of embedding model: dense retrieval, multi-vector retrieval, and sparse retrieval. Multi-Linguality: It can support more than 100 working languages. Multi-Granularity: It is able to process inputs of different granularities, spanning from short sentences to long documents of up to 8192 tokens.

QPC Configurations

Batch Size SEQUENCE LENGTH CORES SoCs QPC URL QPC Size QPC Download Onnx URL Onnx Download Generation Date
1 512 8 1 https://dc00tk1pxen80.cloudfront.net/SDK1.20.4/BAAI/bge-m3/BAAI--bge-m3_qpc_8cores_1bs_512seqlen_1devices_sdk1_20_4.tar.gz 1.1GB Download https://dc00tk1pxen80.cloudfront.net/SDK1.20.4/BAAI/bge-m3/BAAI--bge-m3_ONNX_sdk1_20_4.tar.gz Download 26-June-2026