Frequently asked questions
What is a voice over creator?
A voice over creator is a tool that converts text into speech. Our API generates audio files from text inputs, allowing HR teams to create voice-overs for e-learning modules without hiring talent. It is a cloud-based TTS API that delivers professional audio quickly and scalably for internal communications.
How does AI voice generation work?
Our API uses licensed recordings from LibriTTS-R and LibriVox. You send text via POST request, and the system synthesizes audio using models like tts-1 or tts-1-hd. The process converts up to 4096 characters per request into audio files, which are returned directly in the response for immediate use in your training content.
Is AI voice natural sounding?
Yes, our voices are based on licensed recordings from public-domain audiobook readings. They offer warm, clear, and natural tones suitable for corporate training. While generated audio carries an inaudible synthetic watermark, the output is designed to sound professional and clear for e-learning modules and internal communications without sounding robotic.
Can AI voices mimic humans?
Our voices are based on licensed recordings from LibriTTS-R and LibriVox. They do not mimic specific real people or celebrities. Each voice preset, like Aria or Felix, offers a distinct tone. We do not offer custom voice cloning, so you cannot upload a specific human voice to mimic them exactly.
What are the limitations of AI voice?
Text is limited to 4096 characters per request. We do not support streaming, SSML, or custom pronunciation lexicons. There is no voice cloning or emotion tagging. Audio generation may queue under load. We do not offer an SLA, uptime guarantee, or phone support. Only 23 languages are supported.
How is AI voice used in training?
HR agencies use our API to generate scalable voice-overs for e-learning modules. You convert any text into professional audio in minutes. This allows training departments to produce consistent, natural-sounding narrations for internal communications without the delay of hiring voice talent, ensuring content is ready for immediate deployment.
Is AI voice over better than human?
It depends on your needs. Our API offers scalability and speed, generating audio in minutes rather than days. It provides consistent quality across 8 presets. However, it lacks human nuance like custom emotion tags. It is ideal for high-volume corporate training where consistency and speed are prioritized over unique human performance.
What languages does AI support?
We support 23 languages: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. Any voice can speak any of these languages. If language is not set, it guesses from the script, defaulting to English.
How accurate is the pronunciation?
Pronunciation is based on licensed recordings from LibriTTS-R and LibriVox. The system handles 23 languages accurately. If the language is not specified, it guesses from the script. For non-English Latin text, you should set the language parameter to ensure accurate pronunciation. We do not offer custom pronunciation lexicons for specific terms.
Can I edit the AI voice?
You can adjust the speed between 0.25 and 4.0. You can choose from 8 voice presets and different audio formats like mp3 or wav. However, you cannot edit the voice’s emotion or tone after generation. We do not support SSML for fine-grained control over pauses or emphasis within the audio.
Is the audio royalty-free?
Yes, the generated audio is returned in the response and not stored. Input text is used only to produce the audio and not for training. You can use the generated files for your corporate training modules. The voices are based on licensed recordings, but the output is yours to use.
How long does generation take?
Generation is roughly real-time, about 25 characters per second. Short phrases take a few seconds. Longer texts are split into parts. Requests are processed sequentially, so under load, a request may wait in a queue. Typical speed is fast enough for efficient production of e-learning content for HR teams.
What is the character limit?
Each request supports up to 4096 characters. For longer texts, you can split them into parts and join the audio files. Spaces and punctuation count as characters. This limit allows for substantial content per request while maintaining generation speed and reliability for your corporate training and internal communication needs.
Does it work with video editing?
Yes, you receive audio files in formats like mp3, wav, opus, flac, or pcm. These files can be easily imported into any video editing software. We do not offer a video editing pipeline, but the generated audio is ready to be synced with your visual content.
Understanding the Technology
Our API converts text into audio files via a simple POST request. It uses licensed recordings from LibriTTS-R and LibriVox to generate synthetic speech that includes an inaudible watermark.
The system supports 23 languages and eight distinct voice presets. You control the output format, speed, and language, ensuring the audio fits your specific training modules.
Text-to-Speech basics
Send text up to 4096 characters via JSON, and receive an audio file. The endpoint mirrors the OpenAI speech API, allowing you to use the official OpenAI SDKs by changing the base URL.
- JSON-based request body
- Supports mp3, wav, opus, flac, pcm
- Speed adjustable from 0.25x to 4.0x
Neural network models
We offer two models: tts-1 and tts-1-hd. Both use the same underlying engine and pricing, delivering clear, natural-sounding voice-overs suitable for e-learning without complex infrastructure.
- tts-1 and tts-1-hd models
- Identical engine performance
- Consistent pricing structure
Voice synthesis process
Eight voice presets range from warm narrators to energetic promo voices. The system guesses language from script but allows explicit ISO code setting for accuracy.
- 8 voice presets available
- Automatic or manual language detection
- Licensed source recordings
Benefits for HR Teams
Turn any training script into professional audio in minutes, not days. Our API processes requests sequentially, with typical generation speeds of roughly 25 characters per second for long texts. Short phrases may take a few seconds, but the throughput allows HR teams to scale content creation without waiting for external voice actors.
Speed of production
Generate voice-overs for e-learning modules rapidly by sending text to the API. With a rate limit of 60 requests per minute per key (10 before the first top-up), you can process batches efficiently. Longer texts are split into 4096-character parts and joined seamlessly, ensuring complete modules are ready quickly.
- Text processed at ~25 characters per second
- 60 requests per minute allowed
- Automatic splitting for long scripts
Consistency across modules
Maintain a uniform brand voice across all your training content. By selecting from eight distinct presets like Aria or Felix, you ensure that every module sounds cohesive. The same voice can be applied across different departments, creating a reliable auditory identity for your organization.
- 8 consistent voice presets
- Uniform tone across all modules
- No variation from different actors
Cost efficiency
Eliminate the high costs of hiring voice talent for every update. At $0.01 per 1,000 characters, generating audio is highly economical. Failed requests are free, and your balance never expires, allowing you to budget precisely for ongoing training needs without monthly subscriptions or seats.
- $0.01 per 1,000 characters
- No monthly plans or seats
- Balance does not expire
Comparison with Human Talent
Corporate training often requires recurring voice-overs for updated modules. While human talent delivers unique nuance, it incurs recurring session fees and scheduling delays for every revision.
Cost comparison
Our API charges $0.01 per 1,000 characters. Generating a 10-minute script typically costs a fraction of a single studio hour, with no minimum spend or monthly commitment.
- Pay only for characters generated
- No session or booking fees
- Zero cost for failed requests
Speed comparison
Human production requires casting, recording, and editing, taking days. Our API generates audio in seconds, allowing immediate iteration on scripts without waiting for talent availability.
Quality comparison
Our voices are based on licensed recordings from LibriVox, offering clear, professional narration suitable for e-learning. While they lack human emotional improvisation, they provide consistent, distraction-free audio ideal for instructional clarity.
Use Cases in Training
Turn static compliance manuals into engaging audio lessons. Send any script to our API and receive professional narration in minutes, allowing HR teams to update training content rapidly without scheduling voice talent.
Compliance training
Convert regulatory documents into clear audio modules. Use voices like Atlas for neutral news-style delivery or Felix for energetic promos. Generate consistent audio for annual updates, ensuring every employee hears the same precise wording.
Product knowledge
Create scalable voice-overs for product manuals and feature guides. Split long scripts into 4096-character chunks and join the audio files seamlessly. This allows training departments to produce dozens of modules quickly, maintaining a uniform brand voice across all materials.
Leadership soft skills
Generate natural-sounding dialogues for soft skills scenarios. Use Luna for storytelling or Milo for conversational tones. These presets help create realistic practice environments for employees, enhancing engagement without the high cost of custom recording sessions.
Technical Requirements
The API requires a standard internet connection to send text payloads and receive audio files. Since synthesis is processed sequentially, longer scripts may queue during high load, but typical generation speed is roughly real-time (about 25 characters per second).
API integration
Integrate by sending a POST request to https://api.playtk8.com/v1/audio/speech with your Bearer key. The request body follows the standard OpenAI shape, allowing you to use the official OpenAI SDKs by updating the base_url to https://api.playtk8.com/v1. You can specify the model (tts-1 or tts-1-hd), voice preset, and output format (mp3, wav, opus, flac, or pcm).
- Use standard JSON bodies with
input,voice, andresponse_formatfields. - Supports speed adjustments from 0.25x to 4.0x for pacing control.
- Rate limit is 60 requests per minute per key.
Script formatting
Submit plain text input with a maximum of 4096 characters per request. For longer training modules, split the text into chunks, generate each segment, and join the resulting audio files. Spaces and punctuation count toward the character limit, so ensure scripts are concise to maximize efficiency.
Internet connection
A stable connection is required to upload text and download the resulting audio file. Failed requests incur no charge, but network interruptions may require retrying the synthesis step for that specific chunk.
Audio Quality Standards
Our TTS engine delivers crisp audio suitable for e-learning modules. Each voice preset, from Aria’s warm narration to Atlas’s neutral news style, ensures clear articulation of technical terms and compliance guidelines.
Clear articulation
The API converts text into audio, ensuring clear articulation of words. This clarity is vital for training materials where misinterpretation of instructions can lead to errors.
- Distinct consonant and vowel sounds
- No mumbling or overlapping words
- Consistent clarity across long texts
Natural pacing
Speed ranges from 0.25x to 4.0x, allowing HR teams to adjust for complex topics. Standard generation takes about 25 characters per second, mimicking a comfortable speaking rate for learners.
- Adjustable speed for comprehension
- Real-time generation on long texts
- Consistent rhythm across modules
Professional tone
Voices like Sage (calm, instructional) and Milo (friendly, conversational) provide appropriate tones for different training types. The synthetic watermark is inaudible, keeping the focus on content rather than the medium.

Language Support
Our API supports 23 languages, allowing HR teams to localize training modules for global teams without hiring regional voice actors. Any voice preset can speak any supported language, ensuring brand consistency across all regions.
Multilingual capabilities
We support 23 languages including Arabic, Chinese, French, German, Hindi, Japanese, Korean, Portuguese, Russian, Spanish, and more. If the language parameter is omitted, the system guesses the language based on the script, defaulting to English for Latin text. For non-English Latin-script content, explicitly setting the language code ensures accurate pronunciation.
- Full support for 23 languages
- Automatic language detection for Latin scripts
- Explicit language code support for precision
Accent variety
Each of the 8 voice presets maintains its unique character—such as the warm narration of Aria or the energetic promo style of Felix—regardless of the language spoken. This allows you to maintain a consistent audio identity across diverse linguistic content.
Dialect handling
While we do not offer specific dialect variants, our model handles standard pronunciations for all supported languages. For languages like Chinese or Spanish, the system uses standard pronunciation rules, ensuring clear comprehension for broad corporate audiences.
- Standard pronunciation for all 23 languages
- No specific regional dialect variants
Integration with LMS
Integrate generated audio directly into your Learning Management System workflows. Since our API returns standard audio files, you can easily attach them to course modules without needing proprietary plugins.
Use the generated MP3 or WAV files to enhance video lessons or standalone modules. This ensures consistent voiceovers across all training materials, keeping your learners engaged with professional-grade audio.
SCORM compatibility
Our output files are standard audio formats, making them fully compatible with any SCORM-compliant LMS. You can embed them as supplementary assets in your course structure without worrying about format restrictions.
Video embedding
Generate voiceovers for existing video content or create audio-only modules. Upload the resulting files directly to your LMS library or embed them via standard HTML5 players within your training courses.
Audio file management
Manage your audio assets like any other digital file. Since we do not store the generated audio, you retain full control over your library, storing files on your own servers or cloud storage for easy retrieval.
Scalability
The API handles corporate training workloads by processing requests sequentially, ensuring stability under load. For large modules, split texts into 4,096-character segments to maintain consistent generation quality and avoid timeouts.
With a standard rate of 60 requests per minute per key, you can efficiently generate hundreds of training assets daily. Higher volumes simply require managing multiple keys or batching requests to stay within limits.
Handling large volumes
Process texts up to 4,096 characters per request. For longer scripts, split the content and join the resulting audio files. This approach ensures reliable delivery even during peak training periods.
- Supports 60 requests per minute per key
- Ideal for generating full course audio libraries
- Prevents queue delays with proper segmentation
Batch processing
Generate multiple voice-overs simultaneously by issuing concurrent API calls. Since requests are processed one after another within the rate limit, batching allows you to maximize throughput without manual intervention.
- No need for complex queue management
- Ideal for daily content updates
- Consistent processing order
Automated workflows
Integrate the API into your LMS or CMS using simple HTTP requests. Trigger audio generation whenever a new training module is published, ensuring consistent voice and speed across all materials.
- Easy integration with existing HR systems
- Automated voice selection
- Consistent speed settings (0.25–4.0x)
Security and Privacy
Your input text is used exclusively for generating the audio file and is never used to train our models. Generated audio is returned in the response and not stored, ensuring your training content remains private.
Data encryption
All API requests and responses are secured via standard HTTPS encryption, protecting your scripts and audio files during transit.
Usage tracking
Request metadata, including timestamp, character count, and status, is retained solely for accurate billing and transparency.
No training on data
Unlike some competitors, we do not use your training scripts or generated audio to improve our base models, keeping your proprietary content strictly yours.
Future Trends
We are focusing on core reliability rather than chasing every emerging feature. While the market pushes for instant delivery, we prioritize consistent, high-quality synthesis that integrates seamlessly into your existing training workflows without requiring complex infrastructure changes.
Voice cloning potential
Our service does not offer custom voice cloning for customers, nor do we allow uploading your own voice for cloning. We provide eight distinct, licensed presets instead, ensuring consistent quality without the variability of personalized clones.
Emotion recognition
Our engine does not support emotion tags or dynamic emotional shifts within a single generation. The voices maintain their preset character—such as the warm narration of Aria or the energetic tone of Felix—providing stable, predictable audio for structured corporate learning.
Real-time generation
We do not offer streaming or WebSocket support for real-time partial audio. Requests are processed sequentially, meaning longer texts may wait in a queue. This ensures stable, high-fidelity audio files rather than the lower-quality streaming often associated with real-time playback.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. This ensures that your training materials sound professional and engaging, regardless of the target demographic.
23 languages
Any voice speaks any of these languages.
