Playtest transcripts update: better for multiplayer, kids, laughter and more

Your playtest transcripts now capture more of what players say, and they're noticeably more accurate. The biggest jump is in playtests with children.

That last part matters. Kids' voices are hard to transcribe, and until now words got dropped and speech came out garbled. Because First Findings and our AI analysis read from the transcript, studios testing with younger players couldn't rely on either. That's fixed.

Transcripts also separate speakers now. When more than one person talks during a playtest, you'll see who said what, labeled Speaker 1, Speaker 2 and so on. Useful when a kid is playing with a parent nearby, or in moderated and multiplayer playtests.

You can correct the labels yourself. Rename a speaker to the real person, merge two labels when one voice got split in two, or split them apart when our AI lumped two people together.

You'll also see reactions like laughter and sighs show up in the transcript, so the tone of a moment comes through in the text.

First Findings and AI analysis get sharper at the same time, since they're built on the transcript. Nothing to turn on and nothing to reorder.

Transcripts arrive with your published videos, available now for every studio on every plan.