Introducing the Batch API: Simpler and more efficient for large-scale workloads
Blog post from Voyage AI
Voyage AI has introduced its Batch API, an asynchronous endpoint designed to handle large volumes of requests more efficiently than synchronous alternatives, such as OpenAI's offerings. The Batch API simplifies workflows by managing up to 1 GB file sizes, 100K inputs per batch, and 1 billion tokens per organization, and it provides cost savings of 33%. With a 12-hour completion window, the API is tailored for tasks that do not require real-time processing, such as offline vectorizing for semantic search and large-scale evaluations, allowing users to maximize daytime productivity with overnight runs. It integrates a new Files API for managing input and output files, making it easy to transition from OpenAI's embedding models with minimal code changes.