Embedding / RAG · Updated 2026-10-25

Cohere Embed — what it is actually for

Cohere's Embed 5 family provides Pro embeddings for indexing and Fast embeddings for queries in a shared embedding space. Pricing is per input token, with a free trial key for evaluation.

Short answer: use Cohere Embed when you need enterprise search and rag indexing. It is a usage-based api service from Cohere; the official pricing source is linked below.

Vendor

Cohere

Category

Embedding / RAG

Pricing model

Usage-based API

Free option

A trial API key is available with usage limits and is not intended for production.

What Cohere Embed is used for

• Enterprise search and RAG indexing
• High-volume query embedding
• Multilingual retrieval

Source: official product page and official documentation.

Who Cohere Embed is for

• Teams optimizing retrieval cost
• Enterprise search with high query volume
• Multilingual document retrieval

Source: official documentation.

Who it is not for

• Buyers who need a free production tier
• Simple pipelines where one model is enough

Source: documented plans, limits, and deployment model on the official product page.

Pricing

• Embed 5 Pro text embeddings are listed at US$0.12 per 1M tokens; Embed 5 Fast at US$0.08 per 1M tokens.Source: cohere.com
• Image embeddings are listed at US$0.40 per 1M tokens. Embeddings are billed on input tokens only.Source: cohere.com
• A trial API key allows 1,000 calls per month. Model Vault dedicated instances start at US$2,000/month.Source: cohere.com

Pro and Fast share an embedding space, so a common pattern is to index with Pro and query with Fast.

Full pricing page: https://cohere.com/pricing

Alternatives to Cohere Embed

• OpenAI Embeddings — Text embedding models billed per input token for search and RAG.
• Voyage AI — Embedding and reranking models for RAG, with domain-specific variants.

Official sources

• Official product page
• Official documentation
• Official pricing page