Developers can access Cohere’s Embed v3 models in Elastic’s Inference API with Elasticsearch 8.13 release.
Businesses can create embeddings for data, index in Elastic, and perform vector and hybrid search.
Elastic’s Inference API simplifies vector search, eliminating self-hosting models and enabling single-API-call ingestion pipelines.
Cohere’s embedding models are now supported with int8 compression in Elastic's Inference API.
Developers can reduce storage costs by 75% without compromising search accuracy.

Cohere specializes in creating AI-powered applications for enterprises, offering advanced capabilities in conversational interfaces and search functions through its large language models (LLMs). Their Retrieval Augmented Generation (RAG) toolkit enables accurate, data-informed responses for conversational apps, while their Embed model provides powerful semantic search capabilities across languages. Cohere also improves search result relevancy with their Rerank tool. They provide customizable models with cost-efficient runtime performance, flexible deployment options, and a firm commitment to privacy, ensuring customer data control. Furthermore, Cohere offers educational resources on LLMs through LLM University to help users better understand and utilize their technology.

Cohere

Cohere develops AI language models for enterprise productivity applications.

Cohere Embeddings Now Available Through Elastic’s Inference API

Recent Product News