Frequently asked questions
What is a TTS API?
A TTS API is a cloud service that converts text into audio files. You send text via a simple HTTP request and receive an audio file. It follows the same request shape as the OpenAI speech endpoint, so you can use standard OpenAI SDKs by changing the base URL to https://api.playk999.com/v1. This makes integration easy for developers without needing complex infrastructure.
How is audio generated?
Send a POST request with your text, voice, and format. The server returns the audio file. For texts over 4096 characters, you must split the input yourself into smaller parts and join the resulting audio files. Each request handles up to 4096 characters. Longer texts are processed sequentially, so under load, requests may wait in a queue before generating.
Is the process real-time?
Generation is not strictly real-time. Long texts process at roughly real-time speed (25 characters per second), but short phrases take a few seconds. Requests are processed sequentially, so under heavy load, your request may wait in a queue before synthesis begins. This ensures consistent quality without requiring dedicated streaming infrastructure.
Do I need to install software?
No software installation is required. Since this is a cloud API, you only need an internet connection and the ability to send HTTP requests. You can use the OpenAI SDK library by pointing it to our base URL, or use simple curl commands. Everything happens on our servers, so your local machine only handles the API calls and receives the audio.
Can I use it offline?
No, you cannot use it offline. The API is cloud-based, meaning all text processing and audio synthesis occur on our servers. You need an active internet connection to send your text and receive the resulting audio file. There is no local engine or offline mode available for this service.
What is a voice preset?
A voice preset is one of eight distinct voice styles available in our API. These include female voices like Aria and Nova, and male voices like Felix and Milo. Each preset offers a unique tone, such as warm, energetic, or calm, allowing you to match the audio to your specific content needs.
How are languages detected?
If you do not specify a language, the system guesses it from the script, defaulting to English for Latin text. For non-English Latin-script text, you should explicitly set the language parameter. We support 23 languages, including Arabic, Chinese, and Spanish, ensuring accurate pronunciation across diverse linguistic inputs.
Is the audio copyrighted?
The voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and LibriVox public-domain audiobooks. The generated audio files themselves are yours to use. Each file contains an inaudible watermark marking it as synthetic speech, but this does not restrict your ability to use the audio in your projects or products.
Can I modify the audio?
Yes, you can modify the generated audio files as you wish. The API returns standard audio formats like MP3, WAV, Opus, FLAC, or PCM. You can edit, remix, or process these files using any standard audio software. The only restriction is that the synthetic watermark remains embedded within the audio data.
What is a watermark?
A watermark is an inaudible marker embedded in every generated audio file to indicate it is synthetic speech. It helps identify the audio as AI-generated without affecting the listening experience. Our system adds this automatically during synthesis, ensuring every file carries a unique synthetic signature.
Do I need a developer?
While a developer can help with integration, you do not strictly need one. Our API mimics the OpenAI speech endpoint, making it easy to use with existing tools. You can use simple curl examples or standard SDKs. Many no-code platforms can also integrate with HTTP APIs, allowing non-developers to connect the service.
Can I use it on my website?
Yes, you can integrate the API into your website to generate audio dynamically. Send text from your frontend or backend to our endpoint and receive audio files to embed or play. This is ideal for creating accessible content, narrating blog posts, or adding audio features to your web applications without managing voice servers.
How long does generation take?
Generation speed depends on text length and server load. Long texts process at roughly real-time speed, while short phrases take a few seconds. Requests are processed sequentially, so high traffic may cause slight delays. Typically, you can expect about 25 characters of text to be generated per second under normal conditions.
Is the API secure?
Yes. We use standard Bearer token authentication. Each account has one active key, revocable instantly. Input text is used only for synthesis, not for training. We keep request metadata for billing but do not store generated audio files. Support is by email only, covering all inquiries including security questions.
Understanding the Basics
Our API converts written text into natural-sounding audio files through a simple cloud process. You provide the text, and we handle the synthesis, delivering high-quality audio ready for your application.
Text Input
Send plain text up to 4,096 characters per request. The system processes your input directly, handling spaces and punctuation as counted characters.
- Supports 23 languages including English, Spanish, and Japanese
- Automatic language detection for Latin script (defaults to English)
Audio Output
Receive synthesized audio files in formats like MP3, WAV, OPUS, FLAC, or PCM. Each file contains an inaudible watermark to identify it as synthetic speech.
Cloud Processing
All computation happens on our servers. You send a request via our API, and we return the audio file without requiring local processing power on your device.
- Real-time generation for short phrases
- Queue processing for longer texts
The Request-Response Cycle
Converting text to audio follows a straightforward request-response pattern designed for simplicity. You send a POST request to the API endpoint with your text and configuration settings in a JSON body. The server processes this input and returns the generated audio file directly in the response, ready for immediate use in your application.
POST Request
Send a POST request to https://api.playk999.com/v1/audio/speech. This endpoint accepts your text and voice preferences. Use the standard OpenAI SDK or any HTTP client by setting the base URL to https://api.playk999.com/v1.
- Use POST method
- Target: /v1/audio/speech
- Include Bearer token in header
JSON Payload
Structure your request as a JSON object. Include required fields like model, input, and voice. Optional parameters allow you to adjust speed or specify a language. The API validates the payload and returns an error if fields are missing or invalid.
- Required: model, input, voice
- Optional: speed, language
- Max 4096 characters per input
Audio File Return
Upon successful synthesis, the API returns the audio file in the requested format (mp3, wav, opus, flac, or pcm). Failed requests return an error code without charging your balance. The audio is not stored server-side, ensuring privacy for your indie projects.
- Formats: mp3, wav, opus, flac, pcm
- No server storage
- Failed requests are free

Voices and Languages
Choose from eight distinct presets like Aria, Nova, Luna, and Sage for female voices, or Felix, Milo, Atlas, and Orion for male voices. Each voice has a specific character, from warm narration to energetic promos, all generated from licensed recordings and marked with an inaudible synthetic watermark.
Our API supports 23 languages, including English, Spanish, French, German, Japanese, and Hindi. Any voice can speak any of these languages, allowing you to mix and match without needing separate voice assets for different regions or markets.
8 Preset Voices
Select from eight curated presets. Female options include Aria (warm narrator), Nova (upbeat), Luna (storytelling), and Sage (instructional). Male options feature Felix (energetic promo), Milo (friendly), Atlas (neutral news), and Orion (deep documentary).
All voices are based on licensed recordings from LibriTTS-R and LibriVox. Every audio file includes an inaudible watermark to identify it as synthetic speech.
- Aria: Warm, clear narrator
- Nova: Bright, upbeat
- Luna: Soft, storytelling
- Sage: Calm, instructional
- Felix: Energetic, promo
- Milo: Friendly, conversational
- Atlas: Neutral, news
- Orion: Deep, documentary
23 Supported Languages
We support 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, giving you flexibility in character selection regardless of the target language.
Language Detection
If you do not specify a language, the system guesses based on the script. Latin script defaults to English.
For non-English Latin-script text, always set the language parameter to ensure accurate pronunciation.
Note: Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, and Bengali are not supported.
Audio Formats Explained
Our API returns audio in multiple formats to suit different project needs. You can choose between lossy compression for smaller file sizes or lossless quality for professional use, depending on your application requirements.
MP3 for Web
MP3 is the default choice for most web applications. It offers a good balance between file size and audio quality, making it ideal for embedding directly into websites or mobile apps where bandwidth matters.
- Small file sizes for faster loading
WAV for Quality
Choose WAV when you need uncompressed, high-fidelity audio. This format is perfect for professional editing, voiceovers, or any scenario where preserving every detail of the synthesized voice is critical for the final product.
- Lossless quality
OPUS and FLAC
OPUS provides efficient compression, ideal for modern applications needing small files without significant quality loss. FLAC offers lossless compression, preserving full audio quality while reducing file size compared to WAV, suitable for archival purposes.
- OPUS: Efficient compression
Speed and Tone
Control the pacing of your audio with the speed parameter, which accepts values from 0.25 to 4.0. This allows you to slow down complex instructions for clarity or speed up promotional content. Note that synthesis is not instant; requests are processed sequentially, so short phrases may take a few seconds to generate, while longer texts generally process at roughly real-time speed (about 25 characters per second).
Adjustable Speed
You can fine-tune how fast the voice speaks by adjusting the speed field in your request. Values range from 0.25 (very slow) to 4.0 (very fast). This is useful for making dense information easier to digest or adding energy to marketing clips. Keep in mind that the API processes requests one after another, so under load, your request may wait in a queue before synthesis begins.
Voice Characteristics
Choose from eight distinct presets, each designed for specific tones. Female voices include Aria (warm narrator), Nova (upbeat), Luna (soft storyteller), and Sage (calm instructor). Male options include Felix (energetic promo), Milo (friendly), Atlas (neutral news), and Orion (deep documentary). These voices are based on licensed recordings and deliver consistent, natural-sounding results without requiring complex engineering.
Natural Flow
Each voice provides a consistent tonal quality that fits your project’s style. Aria offers clear narration, while Orion adds depth for documentaries. The natural flow comes from the underlying engine, ensuring smooth transitions without the need for manual pacing adjustments. Simply select the voice that matches your content’s mood and let the API handle the delivery.
Technical Requirements
To use the Voxable API, you only need a stable internet connection and a valid API key. The service is cloud-based, so no local software or SDK installation is required. You can send requests from any device or environment that supports standard HTTP calls.
Internet Connection
Since the processing happens on our servers, your local machine only needs to send the text and receive the audio file. Ensure your network allows outbound HTTPS requests to api.playk999.com. There are no special port configurations needed.
API Key
Authentication is handled via a Bearer token. Include Authorization: Bearer tts-... in your request headers. You can generate or rotate keys directly from your dashboard. If you lose your key, issue a new one instantly; the old one stops immediately.
JSON Format
Send your request as a JSON object in the POST body. Required fields are model, input, and voice. The response_format field is optional; if omitted, the API defaults to MP3. Ensure your text does not exceed 4096 characters per request.
Use Cases in Apps
Our API helps indie developers and no-code creators turn text into natural-sounding audio without writing complex code. You can integrate voice directly into your projects using simple API calls.
Accessibility Features
Generate audio for articles, blogs, or educational content to support users with visual impairments or reading difficulties. With 23 supported languages and 8 distinct voices, you can create inclusive experiences for a global audience.
- Support for 23 languages including English, Spanish, and Japanese
- Clear, natural voices suitable for long-form content
Game Narration
Add dynamic narration to games without recording expensive studio sessions. Use MP3 for web-based games or WAV/FLAC for high-quality console assets where file size is less critical than lossless fidelity.
Split longer scripts into 4096-character chunks to handle full game dialogues. The API returns audio files that you can seamlessly integrate into your game engine.
- Use MP3 for web games to save bandwidth
- Use WAV or FLAC for high-quality console assets
Voice Assistants
Create responsive voice interactions for no-code apps or simple AI bots. Adjust speed from 0.25x to 4.0x to match the pacing of your application.
Use the OpenAI SDK or plain HTTP to send text and receive audio instantly. This allows for quick prototyping and deployment without managing complex infrastructure.
- Adjustable speed from 0.25x to 4.0x
- Simple integration with OpenAI SDK or HTTP

Limitations to Know
Our API is designed for simplicity and predictable scaling, but understanding its boundaries helps you build robust integrations without surprise errors. Knowing these constraints upfront saves debugging time when you deploy your audio features.
4096 Character Limit
Each API request accepts a maximum of 4,096 characters. This limit includes spaces and punctuation. For longer scripts, split the text into chunks and concatenate the resulting audio files.
No SSML Support
We do not support Speech Synthesis Markup Language (SSML). You cannot use tags to control pauses, emphasis, or pronunciation. The engine relies on natural language processing to determine flow.
No Real-Time Streaming
Audio is generated as a complete file, not streamed in real-time. You receive the full audio data once synthesis finishes. This approach ensures high quality but means you cannot hear partial output during generation.
Security and Privacy
Our API uses standard Bearer token authentication to secure your requests. You simply include your unique API key in the Authorization: Bearer header of every call. This ensures that only your account can consume your character balance.
Bearer Token Auth
Each account holds one active key. If you need to rotate credentials, issuing a new key immediately invalidates the old one, preventing unauthorized use during transitions.
- Use
Authorization: Bearer <your-key>in request headers. - Issue a new key anytime; the old one stops instantly.
- One key per account keeps billing simple and secure.
No Data Training
We respect your input privacy. The text you send is used strictly to generate the audio file returned in the response. We do not store your input text for machine learning or training our models.
Metadata Storage
Generated audio files are returned directly to you and are not stored on our servers. We retain only request metadata—such as timestamp, character count, and status—for billing and debugging purposes.
Getting Started
Start by creating an account on our /get-api-key/ page using Google or email. This grants immediate access to a $0.50 free trial, valid for 7 days, allowing you to generate approximately 50,000 characters of audio without payment.
Once registered, you receive a unique API key directly in your dashboard. This key authenticates your requests, and you can issue a new one at any time to invalidate the old one instantly.
Sign Up
Visit the sign-up page and choose Google or email authentication. Accounts are per site, ensuring your credits and usage remain isolated to this storefront.
Get Key
After logging in, navigate to your dashboard to view your active API key. Copy this string; it starts with 'tts-' and is required for all API calls.
First Request
Use the provided curl example in your dashboard. Send a POST request to the API base URL with your text, chosen voice, and model to receive your first audio file.
- Send JSON payload with 'model', 'input', and 'voice'
- Receive MP3, WAV, OPUS, FLAC, or PCM audio
- Check balance and usage stats in the dashboard
Common Mistakes
Developers often overlook the 60 requests per minute rate limit, causing their applications to pause unexpectedly during batch processing. Understanding this constraint helps you design smoother workflows for your projects. Another frequent error is assuming response_format is mandatory; it defaults to MP3 if omitted, so you can skip it for standard web use. Always verify your key’s validity and check for typos in the model name to avoid unnecessary billing for failed attempts.
Ignoring Rate Limits
The API allows 60 requests per minute per key. If you exceed this, requests queue or fail. Design your app to respect this limit, especially for long texts split into multiple parts. This prevents unexpected delays in your user experience.
- Start with 10 requests/minute before your first top-up
- Monitor usage via the dashboard
- Implement exponential backoff in your code
Wrong Format
The response_format field is optional and defaults to MP3. It only fails if you provide an invalid string like 'jpg'. Common formats like WAV, OPUS, FLAC, and PCM are fully supported. Ensure you are using the correct parameter name in your JSON payload.
- Omit for default MP3
- Use 'wav' for high quality
- Use 'opus' for small file size
Missing Voice
If you do not specify a voice, the API may not behave as expected depending on the model. Always explicitly set one of the 8 available voices (e.g., 'nova', 'fable') in your request. This ensures consistent audio output across different calls and environments.
- Available voices: aria, nova, luna, sage, felix, milo, atlas, orion
- Set voice in the JSON body
- Check /v1/voices endpoint for options

Future of TTS
The TTS landscape is evolving rapidly, but our current engine delivers reliable, high-quality results today. We focus on stable, predictable performance rather than chasing every new AI trend, ensuring your projects remain unaffected by underlying model shifts.
AI Advancements
We use proven neural networks rather than experimental prototypes. This means consistent latency and quality without the instability of bleeding-edge models.
- Stable neural engine
- Predictable output quality
Natural Voices
Our eight voices are designed for clarity and ease of listening. As AI improves, we update voices periodically to enhance natural flow without changing your integration.
Integration Trends
We support standard JSON payloads compatible with major SDKs. This ensures easy integration into no-code tools and indie apps without custom adapters.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Indie developers use no-code tools to bring their unique creative game visions to life effortlessly.
23 languages
Any voice speaks any of these languages.
