Embedding Models/Cohere Embed English v3

Cohere Embed English v3 by Cohere — 1024D

English-optimized text and image embedding model. Fast and accurate for classification and retrieval tasks.

At a glance

Modalities

Dimensions

1K

Max tokens

512

Parameters

Price / 1M tokens

Type

Dense

Output types

Single VectorMulti Vector

Language support

en

Benchmarks

MTEB AVG 64.47
MTEB RETRIEVAL 55.00

Details

Release date 2024-01-01
License Proprietary
Model ID cohere-embed-english-v3
Provider Cohere

Tags

text-embeddingenglish-only