Granite Embedding 311M Multilingual R2

Name: Granite Embedding 311M Multilingual R2
Author: Ibm

Ibm🇺🇸 United States

active

Compare with other models →

Context window33K tokens

Version History

r2majorApril 29, 2026

R2 upgrades architecture from XLM-RoBERTa to ModernBERT, extends context from 512 to 32,768 tokens, expands vocabulary to 262K tokens, and adds Matryoshka dimension reduction support. Performance improves by 11.8 points on MTEB Retrieval.

Coverage

model releaseIbm

IBM Releases Granite Embedding 311M R2 With 32K Context, 200+ Language Support

IBM released Granite Embedding 311M Multilingual R2, a 311-million parameter dense embedding model with 32,768-token context length and support for 200+ languages. The model scores 64.0 on Multilingual MTEB Retrieval (18 tasks), an 11.8-point improvement over its predecessor, and ships with ONNX and OpenVINO models for production deployment.

May 6, 2026 · 8:21 AM2 min read

IBM embeddings multilingual

Granite Embedding 311M Multilingual R2 — AI Model Card | TPS