GeckoEmbeddingModel

The Gecko family of models (e.g., `textembedding-gecko`) converts text into dense numerical vectors for high-performance semantic search and AI clustering on Google Cloud's Vertex AI.

GeckoEmbeddingModel is Google Cloud's premier text embedding technology, hosted on Vertex AI. It efficiently transforms text data into high-dimensional, dense numerical vectors, capturing deep semantic meaning. This is critical for applications like building powerful semantic search engines, performing text classification, and identifying data anomalies (outliers). For example, the `textembedding-gecko@001` model supports a maximum input length of 3072 tokens, ensuring high-fidelity representation for large documents. We use this model to power similarity-based retrieval and vector database operations across our core services.

https://cloud.google.com/vertex-ai/docs/generative-ai/embeddings/overview