Cohere Provider

Embedding generation using Cohere's Embed API.

Requirements

• Cohere API key from dashboard.cohere.com/api-keys

Configuration

{ok, State} = barrel_embed:init(#{
    embedder => {cohere, #{
        api_key => <<"...">>,                    % or use env var
        model => <<"embed-english-v3.0">>        % default
    }}
}).

Options

Using Environment Variable

Set COHERE_API_KEY instead of passing in config:

export COHERE_API_KEY=your-api-key

{ok, State} = barrel_embed:init(#{
    embedder => {cohere, #{}}  % uses env var
}).

Supported Models

Input Types

Cohere supports different input types for optimization:

{ok, State} = barrel_embed:init(#{
    embedder => {cohere, #{
        input_type => <<"search_query">>  % for queries
    }}
}).

Example

%% Initialize
{ok, State} = barrel_embed:init(#{
    embedder => {cohere, #{
        api_key => <<"your-api-key">>,
        model => <<"embed-multilingual-v3.0">>
    }}
}).

%% Generate embeddings
{ok, Vec} = barrel_embed:embed(<<"Machine learning is fascinating">>, State).

%% Batch (more efficient for multiple texts)
{ok, Vecs} = barrel_embed:embed_batch([
    <<"Document about AI">>,
    <<"Document about databases">>,
    <<"Document about networking">>
], State).

Pricing

Cohere charges per token. See cohere.com/pricing for current rates.

For high-volume usage:

• Use batch embedding instead of single calls
• Consider embed-english-light-v3.0 for cost savings
• Cache embeddings when possible