GPT-4 fine-tuning is being developed as part of an experimental access program. Preliminary results indicate that GPT-4 fine-tuning requires more work to achieve meaningful improvements over the base model compared to the substantial gains realized with GPT-3.5 fine-tuning.
This model is a fit for customers who want to maximize an established GPT-4 usecase and are willing to pay additional costs and effort to do so.
Our goal with this experimental access program is to learn about quality, safety, and usage. Those learnings may mean these rates adjust if and when the service becomes generally available.
During the experimental access program, GPT-4 fine-tuning is offered at these prices:
|Model||Training||Input usage||Output usage|
|gpt-4||$0.0900 / 1K tokens||$0.0450 / 1K tokens||$0.0900 / 1K tokens|