PLAY HT / VOICE PRODUCTION
Meet the Voice Agent that helps you write, refine and produce voiceovers in one conversation. Bring an idea, a finished script or a recording. Leave with audio you can actually use.
Active subscription or paid trial required. 2 paid credits per request; audio quoted separately.
TAKE / 001
Write a warm podcast opening about calm workdays in 15 words, and prepare the speech quote.
A short creative brief starts the conversation. You can also paste a script and choose to read your exact words.
YOU DIRECT. IT PRODUCES.
A Voice Agent should do more than suggest better words. Play HT connects your conversation to a production workflow, keeping the script, approved settings and finished versions together.
Describe your audience and the kind of narration you want. Paste finished text, add a public article link, or upload or record audio. If an episode brief is missing its source or length, the agent asks you to clarify before drafting.
Check the full script, selected voice, language, Basic/Pro/Max quality and exact generation credits. Ask for changes or approve the quote. Audio transcription has its own approval; a later voiceover needs a separate approval once its text is available.
Play the finished audio and keep working in the same project. Ask for a wording or delivery change while keeping the settings you want. Generating a revised take creates another whole-output version, with a new quote for its actual text.
Download the complete audio when generation finishes. Long source scripts are split into sections and combined into one indexed MP3. You can refresh or close the page while a task runs, then return to the saved project and its results.
A WHOLE EPISODE. NOT JUST A CLIP.
This complete public-domain story adaptation was produced in the live Voice Agent workspace. We supplied the full manuscript and chose exact reading; the recording is the real output, not a one-prompt writing claim.
The approved generation quote was 123 credits, plus one 2-credit request fee. The downloaded file preserves every audio frame from the four sections in order. Refreshing and reopening during generation retained the same task and charges.
Mandarin recording · Anna Su · Pro · tested 30 October 2026
Original example script and synthetic narration. The audio is in Mandarin; the source text is provided in its original language.
MADE FOR SPOKEN WORK
Turn a clear brief into an opening, an introduction or a short story draft. For a complete long episode, bring the finished manuscript or work through shorter sections. Keep revisions in the same conversation instead of losing track of separate takes.
Prepare a script for listening, choose a suitable public voice or an eligible saved private voice, then review the result. Check facts, names and pronunciation before publishing. Qualitative delivery changes can guide the performance; exact timing to a video is a separate edit.
Read a public article or turn an uploaded recording into editable text before creating narration. If a webpage cannot be fetched, paste its body. Transcription and voice generation are separate paid steps, each shown for your approval.
CLEAR COSTS. BEFORE THE TAKE.
Voice Agent uses the same Play HT credits as the voice editor. Each submitted request costs 2 paid credits. Speech generation and transcription are quoted separately, so you can review a proposal without automatically paying to produce its audio.
Basic: 1 credit per 100 characters. Pro: 2. Max: 4. Each section rounds up. Quality access follows your active plan.
Transcription costs 2 paid credits per started minute of verified audio. Uploads are limited to 25 MiB and a 30-minute audio track; workspace storage is 1 GiB per account.
View plans and credit details22 credits
2 + 20One request + one generated section. Up to 2,000 characters in this estimate. Long scripts round each section separately; the workspace quote is authoritative.
It is a chat-based assistant for producing voiceovers: preparing scripts, reading articles, transcribing recordings and creating downloadable speech after approval. This Voice Agent produces audio files for your creative work. It does not answer phone calls or provide a live spoken customer-service conversation.
Yes. Choose exact reading to preserve the words you supply. For revisions, use a saved version as the source and say which settings to keep. Public voices and your eligible private cloned voices are available; creating a new private voice follows the existing voice-cloning consent and plan limits.
A new submitted request uses the 2-credit service fee. Quote approval, playback, downloads and viewing saved versions add no service fee. A new generated version is billed for its actual text. Planning that fails without a useful result refunds its service fee; incomplete failed generation sections return their reserved credits.
Yes. Accepted tasks and results are saved to your account. Refresh or close the tab, then reopen the project to check progress and download the result. Existing completed sections are retained. If a paid provider request has an unknown outcome, it is not silently submitted again.
Workspace media is private to your account and remains until you delete it or your account. Review the script, facts, voice permissions and generated delivery before publishing. Only upload material you may use. Public sharing, video dubbing and timeline editing are outside this workspace. Read the privacy policy for processing details.
THE NEXT TAKE STARTS WITH YOU
Open a private project, tell Voice Agent what you are making, and review the script and quote before creating your audio.
