Effective April 14, 2025, we’ve released new models GPT‑4.1 and GPT‑4.1‑mini to succeed GPT-4o/4o-mini in the API. And we’ve launched a new model in the API: GPT-4o-nano.
The GPT 4.1 models are trained specifically for developers building with the OpenAI API. They outperform GPT-4o and GPT-4o mini across the board, with major gains in coding, instruction following, and long context understanding.
In October 2024, we launched the Vision Fine tuning enabling developers to fine tune GPT-4o model with both images and text. Vision fine tuning costs $25 per million training tokens, $3.75 per million input tokens, and $15 per million output tokens.
Fine-tuning is available for GPT-4o, GPT-4.1, GPT-4o mini, and GPT-4.1 mini to developers on all paid usage tiers (Usage Tiers 1- 5). You can start fine-tuning these models for free by visiting your fine-tuning dashboard, clicking “create”, and selecting the model from the base model drop-down.
You can find current fine tuning prices on our pricing page: platform.openai.com/docs/pricing.