Scene Builder

Choose what avatar sits in each slot, what script tag belongs to that slot, and what BedVibe voice it uses.

Scene title

Visible avatars

Shared stage background 0 ready

Loaded from the built-in backgrounds list.

Dialogue / debate script

Supported starts of line:
[Name] text
[Name|angry] text
(Name|happy) text
Name: text
Name|whisper: text

Ready

Parse a scene to start.

Timeline

Drag clips left or right to set real start time. Overlapping clips now play simultaneously.

Scene Blocks

0 lines

How BedVibe Avatars works

BedVibe Avatars turns a written dialogue into a performed scene with talking characters on a shared stage. You name your speakers, give each one an avatar and a voice, paste the script, and the page generates the spoken audio and plays it back as a scene you can step through line by line — then export the whole thing as a video file.

You can open this page and build a scene without an account. Signing in — the button sits in the corner of the screen — connects your token balance so you can generate the speech. Generating costs one token per syllable; arranging the scene, stepping through lines, replaying audio you have already made and exporting the video cost nothing.

Setting up the stage

Scene title names the piece and travels with the exported scene file.
Visible avatars sets how many characters appear on screen at once.
Shared stage background places every character in the same location, so a debate looks like a debate and a conversation looks like one room rather than a row of cut-outs.

Casting each speaker

Each speaker gets three things: an avatar (the character seen on stage), a voice (the voice that speaks their lines), and a script tag (the name you will type in the dialogue to address them). The script tag is what links your writing to the casting, so a speaker tagged Narrator speaks every line you write under Narrator.

Avatar and voice are chosen independently, which means a character's appearance and their voice are separate decisions. You can recast either one without rewriting the script.

Writing the dialogue

Write your dialogue or debate in the script box, one line per speaker turn, addressing each speaker by the script tag you assigned them. Press Parse Script to split the text into blocks. Parsing is free and generates no audio — it lets you confirm every line has been attributed to the right speaker before you spend anything.

Generating the speech

Generate Current makes audio for the line you are on. Use this while you are still adjusting wording, so a rewrite costs you one line rather than the whole scene.
Generate All makes audio for every block in one pass, once the script is settled.
Each block has its own Generate and Play buttons, so you can work through a scene out of order.

Playing the scene back

Play Scene performs the whole thing from the beginning, with the avatars speaking in turn.
Play Line plays only the current line.
Prev and Next step through the scene one block at a time, which is the fastest way to check timing and delivery.
Stop Audio halts playback immediately.

Exporting

Download Video renders the performed scene — avatars, stage background and generated speech — as a video file you can publish or edit elsewhere.
Download Scene JSON saves the whole scene definition: title, casting, background, script and block structure. Keep it to reload or version a scene rather than rebuilding it from scratch.
Clear Generated Audio discards the audio while keeping the scene, for when you want to regenerate everything with different casting.

What it costs

Speech costs one token per syllable, the same rate as everywhere else on BedVibe. Setting up the stage, casting, parsing, stepping through lines, replaying audio you already generated, downloading the scene JSON and downloading the video do not cost tokens. Only generating a line does, and regenerating one block charges for that block alone.