Table of Contents [expand]
Last updated February 10, 2026
Cohere Embed Multilingual is an advanced embedding model that can convert text and images into dense vector representations across multiple languages. You can compare these resulting vectors to accomplish goals like similarity.
- Model ID:
cohere-embed-multilingual - Regions:
us,eu
When to Use This Model
Cohere Embed Multilingual is ideal for Retrieval-Augmented Generation (RAG) tasks, where you need to search and retrieve relevant documents based on natural-language queries. This model is also useful for building recommendation systems and classification tools that require consistent text embeddings.
Usage
Cohere Embed Multilingual follows our Cohere /v1/embeddings API schema.
To provision access to the model, attach a heroku-inference addon to your app $APP_NAME:
heroku addons:create heroku-inference:standard -a $APP_NAME
You can invoke the model in various ways, see:
- Heroku CLI
aiplugin (heroku ai:models:call) curl- Python
- Ruby
- JavaScript
Multimodal Support
- Supported inputs:
text,image - Supported outputs:
embedding
Rate Limits
- Maximum requests per minute: 500
- Maximum tokens per minute: 800,000
Example curl Requests
Text to Embedding
export EMBEDDING_KEY=$(heroku config:get -a $APP_NAME EMBEDDING_KEY)
export EMBEDDING_URL=$(heroku config:get -a $APP_NAME EMBEDDING_URL)
curl $EMBEDDING_URL/v1/embeddings \
-H "Authorization: Bearer $EMBEDDING_KEY" \
-d @- <<EOF
{
"input": ["Hello, I am a blob of text.", "How's the weather in Portland?"],
"model": "cohere-embed-multilingual",
"input_type": "search_document",
"encoding_format": "raw"
}
EOF