Lyria is an experiment in Creative Space. It uses Google's Lyria 3.5 music model to turn a short meditation script into a single recording, with the spoken voice and the background music generated together. Recordings last up to about three minutes.
Lyria is experimental. The words generally follow the script, but pauses, pacing and timing can vary. The Lyria page starts with Samples, six example recordings you can play before creating your own.
Create a recording
- Open Creative Space in the app navigation, then choose Lyria.
- Under Choose a Meditation, select a completed short meditation from your library.
- Open Meadow prepares a prompt. Review it under Your Prompt, and edit the voice, the music or the script if you like.
- Choose Generate Recording. A recording can take a few minutes.
- Listen under Your Recordings, or choose Download MP3.
Open Meadow checks access and credits when the page opens and again before generating. If something is missing, the page shows what is needed and offers View Account.
Which meditations qualify
Only completed meditations with speech and an estimated duration of 3.5 minutes or less appear in the list. The estimate includes speech, pauses and the meditation's duration and spaciousness settings, rounded to the nearest half minute.
For a meditation longer than three minutes, Open Meadow shortens the pauses toward three minutes. The spoken text and the saved meditation do not change. This cannot guarantee an exact duration, especially when the speech alone is long or the pauses are exact.
If no meditation qualifies, the page offers Create a Short Meditation, which opens the chat. A meditation of about three minutes or less works.
The prompt
Open Meadow writes the prompt from the meditation:
- Opening line: asks for a spoken-word guided meditation, with a duration rounded up to 15 seconds and capped at three minutes.
- Voice: a narrator description chosen from the script's theme, setting and language.
- Background sound: music, nature or place sounds, or nature sounds only.
- Delivery: a fixed line asking Lyria to speak every line, without singing, chanting or humming.
- Lyrics: the spoken text. The title is left out, sections become tags such as
[Section 1], passages have estimated timestamps, and pauses become[instrumental only, no voice]passages.
The voice and music descriptions do not use the meditation's voice or music settings. Choosing an ElevenLabs or custom voice in the player does not make Lyria use that voice. Asking for a longer recording in the prompt does not extend the three-minute limit.
Cost and access
| Item | Credits |
|---|---|
| Each successful recording | 1.5 |
| Preparing the prompt | None |
| Content check | None |
| Listening or downloading again | None |
| A failed generation | None |
Lyria is available on every plan, including Free. It uses included credits first, then extra credits. Every generation needs 1.5 available credits.
Content checks
Before sending a prompt, a quick check confirms that it contains a usable meditation script and applies the same content rules as the Instant Player. It is permissive about formatting, timing notes and music descriptions. Lyria is for guided meditations, not songs or general music.
If the prompt has no meditation script, the page shows "This doesn't appear to be a meditation. Please revise the prompt and try again." For other content issues, it says what to revise. Checks do not use credits.
Saved recordings
Your Recordings lists your Lyria recordings, newest first, each with its own cover image, the source title, the date, a player and Download MP3. Leaving the page does not cancel a recording in progress. It appears the next time you open Lyria.
Lyria recordings are separate from the meditation's regular audio. Deleting the source meditation does not delete its recordings. There is currently no way to delete, share or reopen the prompt of a recording.