OpenAI has intensified the AI arms race by announcing free fine-tuning for its GPT-4o Mini model, just hours after Meta launched its open-source Llama 3.1 model.

While OpenAI had teased the imminent arrival of customization features in last week’s GPT-4o Mini announcement, the timing of this release couldn’t have been more perfect—or more pointed. Just hours after Meta released its Llama 3.1 model, OpenAI fired back with its own offering. Coincidence? Perhaps. But in the high-stakes competition for AI dominance, such precise moves rarely happen by chance.