Fine-Tuning GPT-4o and GPT-4o Mini

Now Available on ALL paid API Tiers

If you’re a software developer using GPT-4o or GPT-4o mini — there’s some good news!

Fine-tuning is now available for all paid usage tiers of OpenAI’s API.

Fine-Tuning GPT-4o and GPT-4o Mini — Created on ImageFX

Before, you had to have a higher-paid tier to fine-tune the models.

I tried making some AI chatbots by fine-tuning with GPT-3.5 Turbo — and the results were ok but this should make a huge improvement.

Fine-tuning on GPT-4o or GPT-4o mini is also free until the end of September — up until 1M daily tokens.

If you don’t have a paid Medium account, you can read for free here

How Much Does It Cost to Fine-Tune OpenAI Models like GPT-4o?

The cost of fine-tuning GPT-4o is normally $25 per million tokens. This information was found on the OpenAI website.

“GPT-4o fine-tuning training costs $25 per million tokens, and inference is $3.75 per million input tokens and $15 per million output tokens.”

Fine-tuning credit until September 23, 2024:

• 1M Free Training Tokens Per Day: For GPT-4o fine-tuning.