mteb-benchmark
1 article tagged with mteb-benchmark
March 31, 2026
model releaseMicrosoft
Microsoft releases Harrier embedding models with 32K context window, achieving 74.3 on MTEB v2
Microsoft released the Harrier-OSS embedding model family, comprising three variants with 270M, 600M, and 27B parameters. The largest model achieves 74.3 on the Multilingual MTEB v2 benchmark. All models support 32,768 max tokens and multilingual inputs across 40+ languages.