> ## Documentation Index > Fetch the complete documentation index at: https://help.descript.com/llms.txt > Use this file to discover all available pages before exploring further. # Speakers > Speakers identify who's talking in your Descript transcript and enable text-to-speech and on-screen avatars. A Speaker is a label that identifies who's talking in your project. At its simplest, it's just a name in your transcript, but it also enables speech generation and avatar features in Descript. * **A [label](/label) on its own** organizes your script. It marks who's talking throughout the transcript. * **A label with an assigned voice** can generate speech with one of [Descript's stock voices](/ai-speech/stock-voices) or a [custom voice clone](/ai-speech/custom-speaker). * **A label with an [avatar](/generative-media/overview)** puts a presenter on screen. ## Project Speakers and drive Speakers Speakers exist at two levels: * **Project Speakers** exist in the project where you created them. * **Drive Speakers** are custom voice clones. Reuse them across projects, and share them with other drive members. They live in the **AI speakers** tab in Drive view, along with any speakers other members have shared with you. ## Working with Speakers * **Add and manage speaker labels** — apply, move, rename, replace, and remove labels in your script * Detect and label speakers in your transcript — let Descript find the voices in a recording * Stock voices * Create a voice clone * Managing and sharing your AI Speakers — rename, share, and delete drive Speakers * Generate text-to-speech audio * Use Regenerate to change what was said * Avatars