Frequently asked questions
What is text-to-speech?
Text-to-speech (TTS) is a technology that converts written text into audible speech. Our API accepts text input and returns an audio file, enabling educators to transform textbooks and lesson plans into audio formats for accessible learning materials.
How does an API work?
You send a POST request to our endpoint with your text and parameters. The API processes the request and returns the generated audio file. You can use standard HTTP libraries or the OpenAI SDK by adjusting the base URL to our domain.
What is a neural TTS?
Our service generates speech from licensed recordings. Voices are based on LibriTTS-R and LibriVox public-domain audiobook readings. Each generated file includes an inaudible watermark to identify it as synthetic speech, ensuring clear origin tracking.
How is audio generated?
Send text via POST request to the /v1/audio/speech endpoint. The system synthesizes the audio using the specified model and voice. For texts over 4096 characters, split the input into parts and join the resulting audio files sequentially.
Can I use it for books?
Yes, you can convert books by splitting text into 4096-character chunks. Each request generates a separate audio file. You must join these files to create the complete book audio. This method ensures compatibility with the API’s character limits.
Is it real-time?
Generation is not instant. Long texts take roughly real-time duration (25 characters per second). Short phrases take a few seconds. Requests are processed sequentially, so high load may cause queuing delays. It is not a streaming or WebSocket service.
What languages are supported?
We support 23 languages including English, Spanish, French, German, Chinese, Japanese, and Arabic. Any voice can speak any supported language. If the language is not specified, it defaults to English for Latin script. Vietnamese and Thai are not supported.
How accurate is the voice?
Voices are based on licensed recordings from LibriVox public-domain audiobooks. The synthesis is accurate to the model’s capabilities. Pronunciation follows standard patterns for the selected language. Custom pronunciation lexicons are not available, so unusual terms may require manual splitting.
Do I need a developer?
Basic usage requires sending HTTP POST requests. You can use the OpenAI SDK or any HTTP client. For simple integration, our documentation provides curl examples. Developers can adjust base URLs and parameters, but no custom SDK is needed for basic use.
Is it expensive for schools?
Pricing is $0.01 per 1,000 characters ($10 per 1 million characters). Failed requests are free. You can start with a $0.50 trial credit. Top-ups are via USDT (TRC-20) with no monthly fees. This pay-as-you-go model suits schools with variable audio generation needs.
Can I listen on mobile?
Yes, the API returns standard audio formats like MP3, WAV, Opus, FLAC, or PCM. These formats are compatible with most mobile devices and media players. You can download the audio files directly from the API response for offline listening.
Is the audio downloadable?
Yes, the API returns the audio file directly in the response. You can save this file to your local storage. The audio is not stored on our servers after generation. You are responsible for downloading and storing the files.
Are there voice options?
We offer eight voice presets: Aria, Nova, Luna, Sage, Felix, Milo, Atlas, and Orion. These include male and female options with different tones. You can select the voice that best fits your educational content style via the API parameters.
How do I choose a voice?
Choose based on your content type. Use Aria for clear narration or Luna for storytelling. For instructional content, Sage or Milo work well. Atlas is neutral for news, while Orion suits documentaries. Select the voice ID in your API request.
Understanding Text-to-Speech Technology
Text-to-speech technology transforms written content into audible audio files, enabling educators to convert textbooks and lesson plans into accessible formats without complex software. This process allows teachers to save time while providing students with an alternative way to consume learning materials.
Converts text to audio files
Users send text via an HTTP request to receive an audio file in the response. The API supports standard formats like MP3, WAV, OPUS, FLAC, and PCM, ensuring compatibility with various educational platforms and devices.
Uses licensed voice recordings
The service utilizes voices based on licensed recordings from LibriTTS-R (CC BY 4.0) and public-domain audiobook readings from LibriVox. Every generated file includes an inaudible watermark to identify it as synthetic speech.
Supports 23 languages
The API supports 23 languages, including English, Spanish, German, French, Chinese, Japanese, Arabic, and Hindi. Any voice can speak any of these languages, with optional language specification for precise control.
How the API Process Works
The API converts text into audio by processing standard HTTP requests. Educators send a POST request to the endpoint with their lesson content, and the system returns a downloadable audio file.
Send text via HTTP request
Use the POST method on https://api.play4447.com/v1/audio/speech. The system accepts text up to 4096 characters per request. For longer textbooks, split the content and join the resulting audio files.
- Use
POSTmethod - Endpoint:
/v1/audio/speech - Max 4096 characters per request
Receive audio file in response
The response contains the generated audio file directly. Supported formats include MP3, WAV, OPUS, FLAC, and PCM. Failed requests are free, ensuring no charge for errors.
- Formats: MP3, WAV, OPUS, FLAC, PCM
- Direct file download
- No charge on failure
Standard JSON request body
Send a JSON body with fields like model, input, voice, and response_format. This structure matches the OpenAI SDK format, allowing you to use existing tools by changing the base_url.
- JSON body structure
- Fields: model, input, voice, format
- Compatible with OpenAI SDKs
Benefits for Content Creators
Text-to-speech APIs transform static educational materials into accessible audio, allowing teachers to distribute lesson plans and textbooks in multiple formats simultaneously. This flexibility ensures that diverse learning needs are met without requiring complex software setups or streaming infrastructure, as the API returns complete audio files rather than streaming partial audio.
Accessibility for users
By converting textbooks and reading materials into audio files, educators make content available to students with visual impairments or reading difficulties. The generated audio files are standard formats like MP3 or WAV, ensuring compatibility with most learning devices and screen readers.
- Supports 23 languages for global classrooms
- Standard audio formats for easy playback
Engagement through audio
Audio versions of lesson materials help maintain student attention by offering a different sensory experience. Teachers can use these recordings for homework assignments, allowing students to listen while commuting or relaxing, thereby increasing interaction with course content.
Time-saving for creators
Educators save hours of preparation time by automating the creation of audio versions for their materials. Instead of recording each lesson individually, teachers can generate high-quality narration instantly, freeing up more time for direct student interaction and lesson planning.
Voice Selection and Customization
Selecting the right voice is critical for educational clarity. Our API offers eight distinct presets, ranging from warm narrators like Aria and Luna to neutral news readers like Atlas. Each voice is derived from licensed recordings, ensuring consistent quality without the need for complex voice design or cloning. Educators can choose the tone that best suits their lesson plans, from storytelling to instructional guides.
Speed adjustment allows you to control the pacing of audio generation, with a range from 0.25 to 4.0 times normal speed. This flexibility ensures that audio materials match the listening level of your students, whether they need slower, clearer enunciation for language learners or faster delivery for quick reviews.
8 preset voices available
The API provides eight voice presets, each designed for specific educational contexts. Female voices include Aria (warm narrator), Nova (upbeat), Luna (soft storytelling), and Sage (calm instructional). Male options feature Felix (energetic promo), Milo (friendly conversation), Atlas (neutral news), and Orion (deep documentary). These voices are based on licensed recordings from LibriTTS-R and LibriVox, ensuring legal compliance and high audio quality.
Each voice can speak all 23 supported languages, allowing you to maintain a consistent auditory experience across different subjects and languages. The voices carry an inaudible watermark to identify them as synthetic speech, which helps in managing copyright and attribution for your educational content.
- Aria: Warm, clear narrator ideal for literature
- Nova: Bright, upbeat for engaging lessons
- Luna: Soft, storytelling for bedtime or calm reading
- Sage: Calm, instructional for step-by-step guides
- Felix: Energetic for promotional or quick updates
- Milo: Friendly for conversational exercises
- Atlas: Neutral, reliable for news or facts
- Orion: Deep, documentary style for history
Adjust speed from 0.25 to 4.0
The speed parameter allows precise control over the audio output, ranging from 0.25x to 4.0x. This feature is particularly useful for language learners who may need slower speech for better comprehension or for advanced students who prefer faster delivery. Adjusting the speed does not affect the voice’s tone or quality, only the rate of speech.
For long texts, such as textbooks or lesson plans, you can experiment with different speeds to find the optimal listening pace. Slower speeds can help with clarity, while faster speeds can save time when reviewing large volumes of content. The API processes requests sequentially, so under load, you may experience slight delays, but the audio quality remains consistent.
- 0.25x: Very slow, ideal for beginners
- 1.0x: Normal speed for standard reading
- 2.0x: Fast for quick reviews
- 4.0x: Maximum speed for rapid consumption
Choose format like MP3 or WAV
The API supports multiple audio formats, including MP3, WAV, Opus, FLAC, and PCM. MP3 is recommended for general use due to its balance of quality and file size, making it suitable for web embedding and mobile apps. WAV and FLAC offer lossless quality, which is beneficial for archival purposes or high-fidelity educational materials.
PCM provides raw audio data, which can be useful for specific technical integrations or further processing. The choice of format depends on your storage requirements and the playback devices your students will use. All formats are returned directly in the API response, allowing for immediate download and integration into your learning management system (LMS).
- MP3: Balanced quality and size
- WAV: Lossless, high quality
- Opus: Efficient compression for web
- FLAC: Lossless, archival quality
- PCM: Raw audio data
Language Support Details
VoxClass supports 23 languages, including English, Spanish, German, French, Japanese, and Chinese. Any voice can speak any of these languages, making it easy for educators to create multilingual learning materials without managing separate voice profiles.
You can specify the language using an ISO code or rely on auto-detection. If the language field is omitted, the API guesses the language from the script. Latin script defaults to English, so non-English Latin-text must specify the language explicitly.
23 languages supported
Our API covers a broad range of languages essential for global education: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. Languages like Vietnamese, Thai, or Ukrainian are not yet supported.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
Auto-detect or specify language
You can explicitly set the language parameter for precision, especially for non-English Latin script text. Without it, the system defaults to English for Latin characters or attempts script-based detection. Specifying the language ensures consistent pronunciation and avoids misinterpretation of dialects.
This flexibility allows methodologists to standardize audio output across different lesson plans without needing to know the specific voice-language pairing in advance.
Clear pronunciation in multiple tongues
All eight voices are trained to deliver clear, natural-sounding speech in any of the supported languages. This consistency helps learners develop accurate listening skills, whether they are studying a new language or reviewing content in their native tongue.
Voices are based on licensed recordings from LibriTTS-R, ensuring high-quality phonetics without the need for complex SSML or custom lexicons, which are not supported by the API.
Technical Requirements
To use the VoxClass API, your application needs a stable internet connection to send HTTP requests and receive audio responses. The service operates entirely in the cloud, so no local processing power or software installation is required on your end.
Internet connection required
Every request travels from your server or script to api.play4447.com and back. Ensure your network allows outbound HTTPS traffic on port 443.
Bearer token authentication
Access is secured via a Bearer token. Include Authorization: Bearer tts-... in your request headers. Each account holds one active key; generating a new key instantly invalidates the previous one.
- Keys start with
tts- - One active key per account
- Invalid keys return 401 errors
Standard HTTP methods
The API uses standard HTTP methods. Use POST for generating speech and GET for retrieving voice lists or model info. The request body is JSON, compatible with the OpenAI SDK structure.
- POST for audio generation
- GET for voices/models
- JSON request body
Common Use Cases
Text-to-speech APIs transform static educational materials into accessible audio formats, enabling educators to distribute lesson plans and textbooks in multiple languages without complex software. This process supports diverse learning environments by converting text into high-quality audio files suitable for various devices.
By leveraging standard HTTP requests, institutions can easily integrate audio generation into their existing workflows, ensuring that content remains engaging and accessible to a wider audience.
Audiobooks for literature
Teachers can convert textbook chapters or literature readings into audio files using the API's 8 preset voices. Each request handles up to 4096 characters, allowing for precise control over pacing and tone. The resulting MP3 or WAV files can be distributed to students for independent study, enhancing retention through auditory learning.
- Convert textbook chapters into audio for independent study
- Use preset voices like Aria or Atlas for clear narration
- Supports multiple languages including English, Spanish, and French
Language learning exercises
The API supports 23 languages, making it ideal for language education. Educators can generate audio exercises with specific voice presets to demonstrate pronunciation and intonation. Students can listen to generated content in target languages, reinforcing listening skills without relying on external recording devices.
- Generate audio in 23 supported languages
- Use voice presets to demonstrate pronunciation
- Create custom listening exercises for students
Accessibility tools
For students with visual impairments or reading difficulties, the API provides an alternative way to access course materials. By converting text to speech, educators ensure that all students can engage with the content equally. The generated audio files are compatible with standard accessibility tools and screen readers.
- Support visual impairments and reading difficulties
- Ensure equal access to course materials
- Compatible with standard accessibility tools and screen readers

Integration with Existing Tools
The API mimics the OpenAI speech endpoint structure, allowing educators to integrate TTS into existing educational workflows without rewriting code. By adjusting the base_url to https://api.play4447.com/v1, you can leverage official OpenAI SDKs to convert textbooks or lesson plans into audio files seamlessly.
This compatibility extends to standard HTTP libraries and simple command-line tools, ensuring that teachers and methodologists can automate audio generation for accessible content using familiar programming environments without needing a proprietary SDK.
Works with OpenAI SDKs
Change the base URL to https://api.play4447.com/v1 and use your API key. The request shape matches OpenAI’s standard, so existing scripts for audio synthesis work immediately.
This allows methodologists to integrate TTS into LMS platforms or custom teacher dashboards using libraries they already know, reducing setup time for converting learning materials.
- Uses standard
POST /v1/audio/speechendpoint - Compatible with official OpenAI SDKs
- No proprietary client library required
Easy curl commands
Test integration instantly with a single command. Replace YOUR_API_KEY and text with your content.
This method works for any programming language with HTTP capabilities, making it ideal for quick prototypes or automated scripts for generating audio for students.
- `curl -X POST https://api.play4447.com/v1/audio/speech -H 'Authorization: Bearer YOUR_API_KEY' -d '{"model":"tts-1","input":"Hello","voice":"nova","response_format":"mp3"}'`
Compatible with standard libraries
Use any standard HTTP library (e.g., Python’s requests, JavaScript’s fetch) to send JSON payloads. The API returns audio files directly in the response body.
This flexibility supports various educational tech stacks, from simple web apps to complex backend systems processing large volumes of text for audio conversion.
Data Privacy Considerations
When you send text to the VoxClass API, the input is processed solely to generate the audio file. We do not use your lesson plans or textbook content to train or improve our underlying models, ensuring your educational materials remain distinct from our training data.
Once the audio is delivered in the response, the generated file is not stored on our servers. This ensures that your converted audio files are transient, reducing long-term data retention risks for your classroom resources.
Text not used for training
Your input text is used exclusively for the immediate production of the audio file. It is not ingested into our model training pipeline, preserving the integrity of your specific educational content.
Audio not stored after delivery
The generated audio file is returned directly in the HTTP response and is not archived on our servers. This minimizes data exposure and ensures that your converted lessons remain under your control.
Metadata kept for billing
We retain minimal request metadata, including timestamp, character count, and status, strictly for accurate billing and usage tracking. This data is not used for content analysis or model improvement.
- Request time
- Character count
- Request status
Performance and Speed
Synthesis speed averages roughly one second of audio per 25 characters of input text. For standard lesson plans or textbook chapters, this means generation occurs at near real-time, allowing educators to preview content quickly without significant delays.
Real-time generation for long texts
Longer texts are processed sequentially by splitting into 4,096-character chunks and joining the resulting audio files. This ensures that even dense educational materials generate audio at a consistent pace, mirroring the natural reading speed.
Queue processing under load
Requests are processed one after another. Under heavy load, a request may wait in a queue before synthesis begins. This sequential processing ensures stability but means generation time can extend during peak usage hours.
Consistent latency
Short phrases typically take a few seconds to generate. While latency is generally predictable, the queueing mechanism means that during high-demand periods, users should anticipate slight delays beyond the base synthesis time.
Troubleshooting Common Issues
When integrating the VoxClass API, educators often encounter predictable hurdles. The most frequent issue involves exceeding the 4,096-character limit per request. Texts longer than this must be split into parts and joined manually, as the API does not handle long-form synthesis in a single call. Another common error stems from invalid or expired API keys, which halt generation immediately.
The service processes requests sequentially, meaning under heavy load, short phrases may experience slight delays while waiting in queue. Understanding these constraints helps teachers maintain smooth workflows when converting lesson plans or textbooks into audio files without unexpected interruptions.
Check character limits
Ensure your input text does not exceed 4,096 characters per request. If your textbook chapter is longer, split it into segments. The API returns separate audio files for each segment, which you must join manually.
Verify API key validity
Confirm your Bearer token is active. If you issued a new key, the old one stops immediately. Check your balance in the dashboard to ensure sufficient funds for the requested synthesis.
Review error messages
The API returns clear status codes. If synthesis fails, the charge is returned. Check for rate limit errors (60 requests/minute) or invalid voice codes. Consult the documentation for specific HTTP status codes.
Future of TTS
As educational content shifts toward multimodal formats, the demand for reliable audio conversion continues to grow. This API provides a straightforward solution for converting textbooks and lesson plans into accessible audio files without complex infrastructure.
Increasing demand for audio
Educators are increasingly seeking efficient ways to distribute learning materials. By offering a simple API endpoint, we support the transition from static text to dynamic audio, ensuring students can access content on various devices.
Integration with LMS platforms
While we do not offer a dedicated WordPress plugin, mobile SDK, or browser extension, our standard HTTP requests allow easy integration with most Learning Management Systems. Educators can automate the generation of audio versions for course readings.
Enhanced user engagement
Audio versions of textbooks support diverse learning styles, particularly for students with visual impairments or dyslexia. The consistent quality of our 8 preset voices ensures that auditory content remains engaging and clear for long-form reading.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Empowering educators with adaptable frameworks to nurture critical thinking and foster genuine curiosity in every learner.
23 languages
Any voice speaks any of these languages.
Sources and further reading
- Wikipedia
- TRON developer hub: TRC-20
- LibriVox public-domain audiobooks
- openai-node on GitHub
- LibriTTS-R corpus (OpenSLR 141)
- ISO 639 language codes
- LibriTTS-R paper (arXiv 2305.18802)
- Python documentation
- Node.js documentation
- W3C WAI: making audio and video accessible
- openai-python on GitHub
- MDN: Web Speech API
