All guides

Publishing Workflow

EPUB or Finished Audio? Choose the Right Spotify Publishing Route

Spotify's manuscript-to-audio tools and finished-audio uploads have different limits and review steps. Choose the route before starting a project, especially when your audio already exists.

5 min readFrom our operating notebook
Separate paths from a manuscript and finished audio files converge on a single audiobook with headphones.
Original editorial illustration

Start with what you have, not the first upload button

Our Google-generated audiobook workflow has a clear handoff: create and download the audio, then prepare that finished package for Spotify. An EPUB used to create the narration and the resulting audio files are different production assets. Confusing those stages can send a finished book back through an unnecessary second narration workflow.

Spotify now documents both manuscript-to-audio creation and uploads of existing audio. The important decision is not simply whether a book uses a digital voice. It is whether Spotify needs to create the narration or whether you are delivering narration that already exists.

This guide is a route-selection checklist, not a claim that we have personally tested every feature in Spotify's creation-tool beta. The operational recommendations draw on the existing Google-to-Spotify handoff; platform requirements come from the official sources below.

Route A: You have an EPUB and need narration

Spotify's Audiobook Creation Tools turn an EPUB manuscript into narration using an ElevenLabs digital voice. The documented sequence includes choosing a voice, reviewing pronunciation, listening to chapter previews, and submitting the result for publication. Spotify identifies these audiobooks as digitally narrated.

This route starts with text, not with your already-mastered MP3 files. Choose it when you intentionally want to create a new audio rendition from the manuscript and can complete the tool's review process. Do not assume that selecting it imports the narration choices or corrections made in a different service.

Before beginning, identify the manuscript revision you intend to narrate. Store a stable copy and a brief list of terms needing attention, such as names, acronyms, unusual place names, and words with more than one pronunciation.

Check access before planning the production schedule

Spotify says the creation tools are available to eligible accounts during beta. Access requires a Spotify for Authors account and a completed payment profile registered in a supported country or region. Its access guide says payment-profile approval may take up to five business days.

The overview currently lists English and Spanish and requires agreement to the Audiobook Creation Opt In Terms of Use. Check the current eligibility guidance and terms for your own account rather than assuming that an ordinary Spotify login provides access.

Our recommendation is to verify the route before committing a release date. Keep an access problem separate from a manuscript problem: converting the file again will not resolve an account that is still awaiting approval.

The five-project limit is not an upload limit

This is the distinction most likely to affect a high-volume catalog. Spotify's creation-limit documentation says each account can create up to five audiobooks from text. Drafts and published projects both count. Once used, a slot is permanent; deleting a draft or unpublishing a title does not restore it.

The same documentation allows up to four generations per project, including regeneration during editing or after publication. Manuscripts must contain no more than 500,000 characters. Those are creation-tool limits, not a five-book ceiling on all audiobooks delivered to Spotify.

Treat a new creation project as a deliberate allocation, not a disposable test. Decide which manuscript deserves the slot, prepare pronunciation and voice decisions before generation, and record remaining capacity. Do not plan around deleting experiments to recover slots.

Route B: You already have the finished audio

Spotify's standard upload guidance accepts MP3, WAV, or FLAC audio. Publishers and self-published authors need the Upload audiobook permission, and a completed payment profile is required before publishing. The same guidance says uploading is free and does not impose a limit on the number of audiobooks uploaded.

That is the appropriate starting point to evaluate when a Google-generated package is already downloaded and ready for review. Keep production and delivery separate: check the audio package you actually intend to publish rather than creating a second narration just because the original EPUB is available.

Spotify's digital-voice guidance accepts providers including Google Play Books and ElevenLabs. When uploading digitally narrated audio, it instructs publishers to select the digital-voice narration option so listeners receive the disclosure. An externally generated voice does not turn this into a Spotify manuscript-creation project.

  • Confirm that the finished audio belongs to the intended manuscript revision.
  • Use the package preflight for chapter order, completeness, artwork, and metadata.
  • Record the narration provider and the disclosure choice with the release record.
  • Verify your rights and the relevant provider terms; file possession alone is not a rights check.

Make one route decision per intended release

A small routing record prevents two operators from starting different workflows for the same book. Give the intended release one record containing its title, distinguishing edition information, manuscript revision, existing audio location if any, chosen route, and current status.

An illustrative decision: if a title has reviewed Google-generated chapter files, route it to finished-audio preflight. If another title has only an approved EPUB, evaluate creation-tool access and capacity before opening a new project. These are examples, not claims about individual titles in our catalog.

Do not launch both paths as insurance against a delay. First check whether a draft or submitted version already exists. If the route needs to change, resolve the existing work deliberately instead of leaving two independent submissions competing to become the intended release.

  1. Identify the exact manuscript and intended audio edition.
  2. Determine whether usable, reviewed narration already exists.
  3. Check route-specific access and limits.
  4. Assign one owner and one route in the release record.
  5. Proceed only after the input package and review responsibilities are clear.

Submission means different work is still pending

For creation-tool projects, Spotify says Submit and Publish starts background audio creation. Once processing completes, publication is automatic; most titles publish within 72 hours. That is not a promise of an additional approval stop after generation.

For finished-audio uploads, Spotify describes a submitted title briefly appearing in Drafts, possible catalog absence during processing for up to 24 hours, and listener availability within three business days. Do not treat a temporary absence as permission to start another upload.

Keep submitted, processing, and publicly verified as separate states. Check the resulting public audiobook before closing the release record, and investigate the specific status if processing fails. The practical payoff is fewer unnecessary projects, fewer duplicate submissions, and a clearer answer to what still needs doing.

Sources and further reading

Official documentation checked for this guide. Platform policies and reporting details can change.