OpenAI Expands its Custom Model Program

OpenAI is expanding its Custom Model Program, aiming to assist business clients in creating customized AI models tailored to specific needs and applications.

During the DevDay conference last year, the AI startup launched the program and it saw significant interest from clients. The Custom Model program was initially tailored for model optimization purposes for specific fields. Now, the expansion will enhance its performance.

They are also offering custom-trained models, which OpenAI developed using its own tools and models, such as GPT-4.  These models are designed for customers who need to deeply optimize their models or integrate new domain-specific information.

For instance, SK Telecom collaborated with the company to enhance GPT-4’s capabilities regarding telecom-related conversations in Korean, while Harvey focused on customizing a model for legal case analysis.

Open AI believes that customized models will become a standard for organizations. As a result, it seeks to make the implementation of AI in specific industries and businesses easier. Compared with general-purpose models, customized models are often more compact and efficient. This means that they could help reduce the pressure on the company’s model-serving infrastructure.

In addition to the expanded Custom Model program, the tech giant has introduced new optimization features for developers who still use GPT-3.5. These include a dashboard that enables the comparison of models based on their performance and quality. The company is also reinforcing the support for integrations with third-party platforms like Weights & Biases. The team added enhancements to tooling as well. While optimization for GPT-4 has not been discussed, it was released early access during DevDay.

Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Intelligent Tech sections to stay informed and up-to-date with our daily articles.