How to make a two-voice dialogue or interview voiceover

Some scripts aren't a single narrator — an interview, a host-and-guest podcast intro, a dialogue in a training video. A single monotone voice reading both sides makes it hard to follow who's talking. GetMeAudio supports two speakers in one project, each with their own voice, without exporting two files and stitching them together yourself.

Step 1 — Add a second speaker

Open the editor and add a second speaker in the Speakers panel. Give each speaker a name and pick a voice for each — mix Standard and Natural voices freely between speakers isn't supported (both speakers share one quality tier per project), but you can pick two very different-sounding voices within that tier.

Step 2 — Mark who's talking

Tag each line with who says it: `<speaker 1/>` and `<speaker 2/>` by position, or `<speaker "Host"/>` / `<speaker "Guest"/>` by name. The editor splits your script by speaker and renders each tag as a visible chip, so the structure of the conversation is easy to see and edit — this works the same whether you typed the tag yourself or inserted it from the Speakers panel.

Step 3 — Draft for free, per speaker

Play script uses your browser's built-in voices for a free draft — each speaker gets a distinct device voice automatically, so you can check the back-and-forth pacing before spending anything. Select any part of the script and Play selection correctly resolves which speaker that portion belongs to, even mid-turn.

Step 4 — Preview and download as one file

Preview and Download both honor the per-speaker voice assignment. The finished download is a single MP3 with both speakers' lines synthesized in their own voice and stitched together in order — no manual editing, no separate exports to merge.

Where this fits

Two speakers is enough for most real cases — a two-host podcast cold open, a Q&A-style ad, a training video with a narrator and a "character" voice, a mock interview. It's a deliberate v1 scope: one shared quality tier keeps pricing simple, and two speakers cover the common case without the complexity of a full multi-character cast.