Play.ht

Role: You are a Play.ht TTS specialist preparing a blog post for neural narration. Constraints: 1) Produce a single SSML document in US English suitable for a 5-6 minute read (approx. 700-900 words). 2) Use <s>, <break time=.../>, <emphasis level=...>, and <prosody rate=...> for natural pacing and emphasis; avoid raw stage directions. 3) Choose one female US voice (name the Play.ht voice). Output format: Provide only the complete SSML block, followed by a one-line note with total word count and chosen voice. Example: include a calm pause before the conclusion using <break time="700ms"/>.

Role: You are a Play.ht voice scriptwriter creating a high-conversion 30-second product voiceover. Constraints: 1) Final spoken duration must be 28-32 seconds. 2) Include two distinct CTAs (first mid-script, second final). 3) Use a British male voice and SSML for pacing and a single emphasis. Output format: Return a single SSML snippet optimized for Play.ht with estimated duration in seconds, approximate word count, and suggested export filename (kebab-case). Example: <emphasis level="strong">Buy now</emphasis> and a <break time="300ms"/> before the second CTA.

Role: You are a content operations lead producing weekly article audio for the next four weeks. Constraints: 1) Generate 4 entries (one per week): title, 2-3 sentence blurb, target length in minutes, recommended Play.ht voice (name + locale), and an SSML 2-3 sentence excerpt. 2) Provide an export filename pattern and priority ranking for QA. 3) Keep each SSML excerpt under 40 words. Output format: JSON array of 4 objects with keys: week, title, blurb, minutes, voice, ssml_excerpt, filename, priority. Example: week="Week 1".

Role: You are a podcast producer preparing narration for a 15-minute episode titled "Product Launch Playbook." Constraints: 1) Output three labeled segments: Intro (0:00-1:00), Main (1:00-13:00) with two clear ad slots (at ~4:00 and ~9:00, each ~20 seconds), Outro (13:00-15:00). 2) Use a neutral US male voice; include SSML markers for timestamps, ad boundaries, and a 20s ad script for each slot. 3) Provide recommended export filename and suggested RSS episode summary (two sentences). Output format: JSON with keys intro, main, ads (array), outro, filename, rss_summary.

Role: You are an audio engineer designing a Play.ht voice-cloning workflow for commercial narration. Multi-step constraints: 1) Produce a step-by-step checklist covering legal consent, recording specs (mic, sample rate, quiet room), dataset size and diversity, file formats, metadata tagging, and secure upload steps. 2) Provide 6 SSML test lines (short to long) to validate tonal match; include two few-shot example lines demonstrating tonal variety: Example A: "Welcome back-let's get into today's strategy." Example B: "Quick pause. Now the key number: forty-five percent." 3) End with an acceptance metric table (MOS/LSM targets). Output format: Structured checklist, SSML tests, and metric table in plain text.

Role: You are a localization director creating Play.ht-ready audio scripts for a 90-second brand video. Constraints: 1) Produce transcreated scripts for Spanish (LATAM), French (France), German, and Japanese, each adapted for culture and timing to match 90 seconds ±5s. 2) For each language, specify a recommended Play.ht voice (name and locale) and provide an SSML version with pacing adjustments. 3) Provide a fallback English short-form lines file and a sample transcreation example showing the English line and the Spanish adaptation. Output format: JSON mapping language -> {voice, ssml_script, estimated_seconds}.

Compare Play.ht with ElevenLabs, Murf AI, Speechify, Amazon Polly, Google Cloud Text-to-Speech. Choose based on workflow fit, pricing limits, integrations, governance needs and whether the output must be production-ready or only assistive.