cinara

Appearance

A YouTube video ​

The hard part of a channel isn't the idea. It's the eight hours between the idea and the upload. This workflow hands Cinara the parts that don't need you: the voice track, the cutaways, the thumbnail, the music and the subtitles.

1. Record the narration ​

Open Speech and paste your script. Scripts up to 50,000 characters are read in one go and joined.

  • Your own voice. Clone it once in Voice Cloning — it takes a short consent recording from you — and every later video is read in it.
  • A narrator. Mentor is patient and clear, Canyon is low and cinematic, Kindred is conversational.
  • Pronunciation. Add channel names, product names and anything the voice gets wrong to the pronunciation list. It applies to every script from then on.

Speech is 1 credit a character. A 10-minute script is roughly 9,000 characters, so about 9,000 credits.

2. Make the b-roll ​

For every moment without footage, open Video and describe the cutaway at 16:9. Use From an image when a shot should start on something specific, like a diagram you made in Images.

3. Make the thumbnail ​

Open Images, set Shape to 16:9 and How many to 4. Describe the scene and leave empty space for the title text you'll add in your editor. Frame, the default look, renders text accurately if you want words in the image itself.

For a recurring character or host, keep them consistent: choose Edit a picture, add earlier thumbnails as references, and describe the new scene.

4. Add the music bed ​

Open Music, choose Instrumental and describe something that sits under a voice — "gentle ambient pads, soft piano, unobtrusive, 80 bpm". Music you make in Cinara is yours, so it isn't claimed.

5. Subtitles, chapters and the description ​

Open Transcribe and add the finished video.

  • Download SRT subtitles and upload them to YouTube.
  • Use the timestamps and speaker labels to write chapters.
  • Copy the transcript into the description; it's also what makes the video searchable.

6. A second audience ​

Open Dubbing, add the finished video and choose a language. Each speaker gets a matched voice, the timing follows the original, and the music stays. It's 2,000 credits a minute of video for each language.

Cinara · AI voice, music and video

© 2026 Cinara