PLOPS FORGE · VOICE STUDIO · OCTOBER 25, 2026

Turn a long script into a voiceover with Ming 3.0

A practical workflow for narration, explainers, and spoken drafts using Ming 3.0 on Forge.

Open Ming voice studio

A useful narration workflow starts with a short audition. In Forge’s Ming studio, pick James, Stella, or an authorized reference voice, then enter a few sentences from your script.

Direct the performance

Describe delivery separately from the selected voice: “steady and conversational” or “quiet, thoughtful, with gentle emphasis.” A short test helps you judge pace and pronunciation. Keep the seed fixed when comparing wording or direction, while remembering that generated voices can vary.

Expand to the complete script

Paste the longer narration into the same field. Forge counts the chunks and quotes the full credit cost. It processes the chunks in sequence and delivers one combined WAV after they all finish. You do not need to split and join files manually.

Keep working while it runs

The tab title reports completed-chunk progress. Your account’s take history provides the finished audio and download link. Refreshing the page does not submit another take.

Review before publishing

Listen for pronunciation, pacing, and transitions between chunks. Use reference voices only with permission. A generated take is a draft you can review and edit, not a guarantee of a particular performance.

Plan credits and availability

The studio quotes the total before charging, currently at 100 credits per chunk. Cold starts add waiting time, and the service’s monthly compute allowance limits capacity. Check availability before buying a pack.

Model source and documentation: inclusionAI/Ming-3.0-tts. The upstream source is MIT licensed.