OpenAI Enhances Assistants API with New Features for Enterprise Users
OpenAI has recently unveiled a range of new tools and features for its enterprise-grade Assistants API, catering to the needs of large businesses requiring customized AI solutions. These updates aim to provide enhanced support to major clients like Klarna, Morgan Stanley, Oscar, Salesforce, and Wix, among others. The company has introduced two new tools and six new features to its Assistants API suite, along with cost-effective options for businesses looking to leverage AI products.
One of the key tools introduced is Private Link, offering increased security by establishing a direct connection between Microsoft Azure and OpenAI to minimize exposure to the open internet. Additionally, a new tool called Projects has been designed to provide better administrative control by allowing project owners to manage roles, API keys, and access to specific projects.
Within the Assistants API, OpenAI has added several new features, including expanded file_search capabilities, support for parallel queries, improved reranking, and streaming support. The API now includes technical components like vector-store objects and tool_choice parameters while also offering support for fine-tuned GPT 3.5 Turbo models.
To help businesses manage costs effectively, OpenAI has introduced two new methods. Clients with a sustained token-per-minute level can request provisioned throughput for up to a 50% discount, while non-urgent workloads can benefit from batch API requests priced at a 50% discount off shared prices.
Overall, these updates underscore OpenAI’s commitment to supporting enterprises with robust AI solutions tailored to their specific needs. By providing enhanced security, administrative control, and cost-effective options, the company aims to empower businesses to leverage AI technology effectively in their operations.
