OpenAI just dropped some exciting news about their enterprise-grade Assistants API.
They’ve rolled out a bunch of new tools and features to better support their big-name clients like Klarna, Morgan Stanley, Oscar, Salesforce, and Wix. Large businesses and developers who are rapidly scaling on the OpenAI platform will benefit from these updates.
First up, let’s talk about the new tools. OpenAI has introduced “Private Link,” which is all about boosting security for businesses that share their organizational data with the OpenAI cloud.
Private Link creates a direct connection between Microsoft Azure and OpenAI, which minimizes exposure to the open internet. Plus, they’ve added native Multi-Factor Authentication (MFA) to provide an extra layer of security and ensure compliance with access control requirements.
But wait, there’s more! OpenAI has also introduced a tool named “Projects,” providing administrators with enhanced control over their API usage.
With Projects, admins can scope roles and API keys to specific projects, restrict and choose the models they want to make available, and set usage and rate-based limits. Project owners can even create service account API keys, allowing other users to access projects without needing their own API keys.
The new features for the Assistants API. The file_search retrieval feature has gotten a massive upgrade, now supporting up to 10,000 files per assistant (up from just 20 before!).
OpenAI claims that this feature now supports parallel queries through multi-threaded searches and boasts improved reranking and query rewriting. They’ve also added streaming support, which is pretty cool.
The Assistants API now includes technical components like vector-store objects and a tool_choice parameter. And get this: it will even support fine-tuned GPT 3.5 Turbo models.
OpenAI recognizes the importance of budget management for businesses, which is why they have implemented two new cost management methods. First, if you have a sustained level of token-per-minute usage, you can now request access to provisioned throughput and get up to a 50 percent discount.
But what if you have non-urgent workloads? No problem! With batch API requests, OpenAI has you covered. OpenAI offers these requests at a 50% discount from shared prices, guaranteeing results within 24 hours and providing high rate limits. It’s a great way to save some cash without sacrificing performance.
With these new tools and features, OpenAI is clearly committed to supporting their enterprise clients and developers. By enhancing security, providing better administrative control, upgrading the Assistants API, and offering cost-effective options, they’re making it easier than ever for businesses to leverage the power of AI.
So, whether you’re a large enterprise or a developer looking to scale quickly, OpenAI’s got your back. With these new additions to their platform, the possibilities for AI-driven innovation are endless. Get ready to take your projects to the next level with OpenAI’s enterprise-grade Assistants API!
The Information is Taken from Gadgets360 and zdnet
