OpenAI is expanding its Custom Model program to help enterprise customers develop tailored generative AI models for specific use cases, domains, and applications. The program offers assisted fine-tuning and custom-trained models to optimize performance on particular tasks. OpenAI emphasizes the need for organizations to develop personalized models that are personalized to their industry and use cases. The company has seen success with customers like SK Telecom and Harvey, who have fine-tuned models to improve performance in specific areas such as telecom-related conversations and case law. By offering tailored models, OpenAI aims to maintain its revenue growth and alleviate strain on its model serving infrastructure.
| Signal | Change | 10y horizon | Driving force |
|---|---|---|---|
| OpenAI expanding Custom Model | Offering tailored generative AI | More organizations | Meeting specific industry needs and maximizing performance |
| Assisted fine-tuning added | Fine-tuning methods at larger scale | Improved model performance | Enhancing model performance in specific tasks |
| Custom-trained models introduced | Custom models built with OpenAI | Personalized AI models | Organizations seeking domain-specific knowledge |
| Increased demand for custom models | Organizations developing personalized models | Customized AI implementations | Need for more meaningful and specific AI impact |
| Revenue growth potential | Custom model training as consulting work | Increased revenue | Maintaining revenue growth while scaling infrastructure |
| Strain on model serving infrastructure | Tailored models are smaller and more performant | Improved infrastructure efficiency | Addressing compute capacity challenges |
| New model fine-tuning features | Improved quality and performance | Enhanced development tools | Supporting developers and integration with third-party platforms |