I built a pipeline that automatically generates AI commentary and voiceover for gameplay videos — using GPT-4o Vision to analyze frames and TTS to narrate them. Built days after the GPT-4 Vision API launched in November 2023, it is one of the earliest known implementations of this approach.






