The EmbeddingsBuilder provides a fluent interface for generating vector embeddings from text input. Embeddings can be used for semantic search, similarity comparison, and other vector-based operations.
Configuration Methods
model
Set the embedding model to use.
$builder->model('text-embedding-3-small')
$builder->model(new OpenAIEmbeddings('text-embedding-3-large'))
model
string|EmbeddingModel
required
The embedding model identifier string or EmbeddingModel instance
Returns the builder instance for method chaining
$builder->input('The quick brown fox jumps over the lazy dog')
$builder->input(fn() => $article->content)
The text to embed, or a closure that returns the text
Returns the builder instance for method chaining
Callback Methods
onTokenStats
Set a callback for token usage statistics.
$builder->onTokenStats(function(TokenStats $stats) {
echo "Tokens used: {$stats->total}";
})
onTokenStats
Closure(TokenStats): void|null
required
Closure that receives TokenStats with usage information
Returns the builder instance for method chaining
Execution Methods
get
Generate the embeddings for the configured input.
$embedded = $builder->get()
EmbeddedData instance containing the vector embeddings and metadata
Thrown if input or model is not set before calling get()
Usage Example
use Mateffy\Magic;
// Simple usage with facade
$embedded = Magic::embeddings(
input: 'Hello world',
model: new OpenAIEmbeddings('text-embedding-3-small')
)->get();
// Advanced usage with builder
$embedded = Magic::embeddings()
->model('text-embedding-3-large')
->input(fn() => $document->content)
->onTokenStats(function($stats) {
Log::info('Embedding tokens', ['total' => $stats->total]);
})
->get();
// Access the embedding vector
$vector = $embedded->vector; // array of floats