Frequently asked questions
What is the difference between TTS API and voice synthesis?
Voice synthesis is the broad technology of converting text to speech. A TTS API is the specific cloud service that implements this technology, providing a standardized interface for developers to generate audio files programmatically. BotVoice offers a RESTful API that handles the synthesis engine, allowing you to integrate voice capabilities into your applications without managing the underlying infrastructure.
How does a TTS API work?
You send a POST request with JSON containing your text, voice selection, and language to the API endpoint. The server processes the text using licensed voice models and returns the audio file in your chosen format, such as MP3 or WAV. You can adjust parameters like speed and model type directly within the request payload.
Can I use TTS API for real-time voice calls?
The API does not support streaming or WebSocket connections for real-time audio delivery. It returns complete audio files after processing. While requests are processed sequentially, typical generation speed is roughly real-time for long texts, but short phrases may take a few seconds. For live calls, you would need to buffer the full audio file.
What is the latency of a TTS API?
Generation speed is roughly real-time, processing about 25 characters per second for long texts. Short phrases may take a few seconds to generate. Requests are processed one after another, so under load, a request may wait in a queue, increasing latency. This ensures consistent quality but means no guaranteed instant response times.
Do I need an SDK to use a TTS API?
No, an SDK is not required. You can use standard HTTP requests or any HTTP client library. The API endpoint matches the OpenAI speech endpoint shape, so the official OpenAI SDKs work by changing the base URL. This allows easy integration into any programming language that supports HTTP calls.
How many languages can a TTS API support?
The API supports 23 languages, including English, Spanish, German, French, Japanese, Chinese, and Arabic. Any voice can speak any of these languages. If the language is not specified, it is guessed from the script, defaulting to English for Latin text. Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, Bengali and others are not supported.
Is the audio generated by a TTS API human-like?
The voices are based on licensed recordings from LibriTTS-R and LibriVox public-domain audiobook readings. They are natural and clear, designed for practical use in bots and assistants. Each generated file carries an inaudible watermark marking it as synthetic speech, distinguishing it from human recordings without affecting the listening experience.
Can I control the speed of the speech in a TTS API?
Yes, you can control the speech speed by setting the speed parameter in your request. The acceptable range is from 0.25 to 4.0, allowing you to slow down the narration for clarity or speed it up for quick updates. This gives you precise control over the pacing of the generated audio.
What is the character limit for a TTS API request?
Each request can contain up to 4096 characters of text. If your input exceeds this limit, you should split the text into parts of up to 4096 characters, generate audio for each part, and join the resulting audio files. This ensures efficient processing and maintains audio quality across longer texts.
Do TTS APIs support SSML?
No, this API does not support SSML (Speech Synthesis Markup Language). You cannot use SSML tags to control pronunciation, pauses, or emphasis. Instead, you control the output through specific parameters like voice selection, language code, and speed adjustment in the JSON request body.
How do I choose the right voice for my bot?
Choose from eight presets based on your bot's personality. Options include female voices like Aria (warm narrator) and Nova (upbeat), or male voices like Felix (energetic) and Atlas (neutral). Select a voice that matches your brand's tone, such as Sage for instructional content or Orion for documentary-style narration.
Is the data sent to a TTS API stored?
Input text is used only to produce the audio and is not used for training. The generated audio is returned in the response and not stored. Request metadata, such as time, character count, and status, is kept for billing purposes. Your raw text input is not retained after processing.
Can I clone my own voice for a TTS API?
No, custom voice cloning is not offered. You cannot upload your own voice or design a new voice. You must choose from the eight available presets. These voices are based on licensed recordings and are available for immediate use without additional setup or training processes.
What are the common use cases for TTS API in chatbots?
BotVoice is designed for product teams integrating reliable text-to-speech into voice assistants and chatbots. Use cases include providing audio responses in conversational bots, generating voice notifications, and enhancing user experience in voice-enabled applications. The straightforward API and clear pricing help speed up voice UX development without managing infrastructure.
Definition of TTS API
A Text-to-Speech API is a cloud service that converts written text into spoken audio. You send a string of characters via an HTTP request, and the service returns an audio file containing the synthesized speech. This process abstracts the complex neural network inference, allowing your application to focus on user interaction rather than audio generation mechanics.
The system processes your input on remote servers to produce audio. This cloud-based approach ensures that your application does not need to manage heavy computational resources or maintain specialized audio hardware to deliver voice features to end-users.
Text Input to Audio Output
The core function of the API is straightforward: input text becomes output audio. You provide the text content, specify the desired voice and language, and receive a digital audio file. The service handles the conversion of characters into phonemes and then into audible sound waves, delivering the result as a downloadable file like MP3 or WAV.
- Send text up to 4096 characters per request
- Receive audio in formats like MP3, WAV, OPUS, FLAC, or PCM
- Specify voice and speed parameters in the request
Cloud-Based Processing
All processing occurs on our servers, not on your device. When you send a POST request to the endpoint, the text is analyzed by our engine, which generates the audio file. The result is returned directly in the response body, allowing your application to play the audio immediately or store it for later use. This model keeps your client lightweight and scalable.
- No local audio engine required
- Automatic handling of voice synthesis algorithms
- Scales with your request volume
Standardized Endpoints
The API uses a simple RESTful structure with a JSON request body. The endpoint POST https://api.playfk9.com/v1/audio/speech accepts standard parameters like model, input, voice, and response_format. This design allows you to use existing HTTP clients or the official OpenAI SDKs by simply adjusting the base URL, making integration into voice assistants and chatbots quick and reliable.
- POST /v1/audio/speech for generation
- GET /v1/voices to list available options
- JSON body with standard fields like
modelandinput
Role in Voice Assistants
BotVoice enables voice assistants to deliver spoken responses by converting text into audio files via a simple API. Product teams integrate this capability to create conversational interfaces that respond to user prompts with natural-sounding speech, eliminating the need to manage complex audio infrastructure.
While not true streaming, BotVoice generates audio files quickly enough for responsive interactions. The API returns standard formats like MP3 or WAV, allowing assistants to play responses immediately. This direct file delivery ensures compatibility with existing voice assistant frameworks without requiring custom playback logic.
Conversational Interface
Voice assistants rely on clear, human-like speech to maintain natural dialogues. BotVoice provides eight distinct voices, from warm narrators to energetic promo styles, allowing teams to select a tone that matches their brand identity. Each voice is based on licensed recordings, ensuring high quality without licensing headaches.
- Support for 23 languages with automatic detection
- Eight preset voices for consistent branding
- No custom voice cloning required
Response Time
BotVoice generates audio files quickly enough for responsive interactions. Long texts are processed at roughly real-time speed, while short phrases take a few seconds. Requests are processed sequentially, so under load, a request may wait in a queue before synthesis begins.
- Roughly real-time generation for long texts
- Sequential processing ensures consistency
- No streaming or WebSocket support
User Experience Enhancement
By offloading speech synthesis to the API, product teams can focus on conversation logic rather than audio encoding. The API accepts text input and returns an audio file, simplifying integration into chatbots and voice apps. Users receive clear, intelligible speech without buffering delays typical of complex TTS engines.
- Standard JSON request format
- Multiple audio formats (MP3, WAV, OPUS)
- No SSML or phoneme dictionaries
Role in Chatbots
Text-to-speech APIs enable chatbots to deliver audio responses, transforming static text exchanges into dynamic voice conversations. By converting written prompts into natural-sounding speech, developers can create more engaging and accessible user experiences without building complex audio infrastructure from scratch.
This capability is particularly valuable for hands-free interactions or users who prefer listening over reading, allowing chatbots to serve a broader audience effectively.
Audio Chatbots
Audio chatbots leverage TTS APIs to read out responses directly to the user. This is ideal for scenarios where visual attention is limited, such as driving or multitasking. The API handles the conversion from text to audio seamlessly, ensuring consistent voice quality across all interactions.
- Instant text-to-audio conversion for rapid response delivery
- Support for multiple voices to match brand personality
- Standardized API integration simplifies audio output implementation
Voice-First Interfaces
Voice-first interfaces prioritize audio input and output, making TTS a core component. Users interact primarily through speech, requiring the chatbot to respond with clear, natural audio. The API’s ability to handle various languages and voices ensures that voice-first experiences feel personalized and responsive.
- Natural-sounding audio responses for voice-driven UX
- Support for 23 languages to reach global users
- Quick integration into existing voice assistant frameworks
Multimodal Interaction
Multimodal chatbots combine text, audio, and visuals to enrich user interaction. TTS APIs allow these systems to provide audio feedback alongside visual cues, enhancing comprehension and engagement. This approach caters to diverse user preferences, offering flexibility in how information is consumed.
- Synchronized audio and visual responses for enhanced clarity
- Flexible output formats (MP3, WAV, OPUS) for various devices
- Seamless integration with visual chat interfaces
Technical Implementation
Integrating BotVoice into your voice assistant or chatbot requires sending a structured JSON payload to the API endpoint. The request includes the text to be converted, the chosen voice model, and the desired audio format, ensuring precise control over the output generation.
The system processes this input and returns a downloadable audio file. This straightforward exchange allows developers to seamlessly embed synthetic speech into their applications without managing complex infrastructure or encoding details.
JSON Request Body ['Send a POST request to /v1/audio/speech with a JSON body. Include the input text (up to 4096 characters), model (tts-1 or tts-1-hd), voice preset, and response_format (mp3, wav, opus, flac, or pcm). You can also adjust speed between 0.25 and 4.0.', 'Include an Authorization: Bearer header with your API key. If needed, specify the language ISO code to ensure accurate pronunciation, though the system guesses from script if omitted.'] ['model: tts-1 or tts-1-hd', 'input: max 4096 characters', 'voice: 8 preset options', 'response_format: mp3, wav, opus, flac, pcm', 'speed: 0.25–4.0', 'language: optional ISO code']
Audio File Response ['Upon successful synthesis, the API returns the audio file directly in the response body. You can save this file or play it in your user interface. Failed requests incur no charge and return no audio data.', 'The generated audio carries an inaudible watermark marking it as synthetic speech. No audio files are stored server-side after delivery, ensuring minimal data retention for your users.']
SDK Integration ['Use the official OpenAI SDKs by changing the base_url to https://api.playfk9.com/v1. This allows you to leverage existing client libraries for straightforward integration into your codebase.', 'Alternatively, use any standard HTTP client to send the JSON request. This flexibility ensures compatibility with your existing tech stack without requiring a proprietary BotVoice SDK.']
Benefits Over Traditional Services
Integrating BotVoice requires minimal engineering effort compared to managing legacy TTS infrastructure. By supporting the standard OpenAI-compatible request shape, teams can use existing SDKs simply by updating the base URL. This eliminates the need for custom wrappers or complex middleware, allowing product teams to deploy voice features faster with familiar HTTP endpoints and JSON bodies.
The pricing model removes financial risk associated with unused capacity. At $0.01 per 1,000 characters, costs scale linearly with usage. Failed requests are free, and there are no monthly subscriptions, seats, or minimum spend requirements. This pay-as-you-go structure ensures that product teams only pay for the actual audio generated, optimizing budget efficiency.
Simplified Integration
BotVoice uses the same request shape as the OpenAI speech endpoint. Teams can switch by changing the base_url to https://api.playfk9.com/v1 and using their API key. This compatibility with standard tools means no proprietary libraries are needed.
The API returns audio files directly, supporting MP3, WAV, Opus, FLAC, and PCM. This standardization simplifies playback logic across different voice assistant and chatbot platforms without custom codec handling.
- OpenAI SDK compatible
- Standard JSON endpoints
- No custom middleware required
Cost Efficiency
Pricing is $10 per 1 million characters. Spaces and punctuation count toward this limit. Failed synthesis requests are charged zero, ensuring you only pay for successful audio generation.
There are no monthly plans or hidden fees. Accounts start with a $0.50 trial credit. Top-ups via USDT (TRC-20) are processed automatically, providing transparent spending control.
- $0.01 per 1,000 characters
- Free failed requests
- No monthly subscriptions
Scalability
The API handles 60 requests per minute per key, with a queue system for higher volumes. Text is split into parts of up to 4,096 characters and joined, allowing long documents to be processed without manual chunking logic.
Infrastructure is managed entirely in the cloud. Product teams can scale voice features from prototype to production without provisioning servers or managing capacity limits, focusing on user experience instead of backend logistics.
- Automatic text splitting
- Cloud-managed infrastructure
- No server provisioning needed
Voice Selection Strategy
Selecting the right voice is critical for voice assistants and chatbots. BotVoice offers eight distinct presets that allow product teams to align audio output with their brand’s personality. Each voice carries a specific tonal quality, from warm narration to energetic promo styles, enabling consistent user experiences across different interaction points.
This selection process ensures that the synthetic voice feels like a natural extension of the product interface rather than a generic add-on.
Brand Consistency
Consistency builds trust. By locking in a specific voice preset for all user interactions, teams maintain a recognizable auditory identity. BotVoice’s voices are designed to be distinct yet professional, ensuring that the audio output matches the visual branding without requiring complex audio engineering.
Use the same voice for onboarding, help, and confirmation prompts to reinforce brand recognition.
- Stick to one preset for core flows
- Use consistent tone for error messages
Audience Preferences
Different user groups respond differently to vocal tones. For instance, 'Sage' offers a calm, instructional tone, while 'Felix' provides an energetic vibe suitable for promos. Analyzing user feedback helps determine which preset resonates best with your specific demographic.
A/B testing voice presets can reveal which style reduces user confusion or increases engagement in chatbot scenarios.
- Match voice to target demographic
- Test presets with user groups
Voice Diversity
While consistency is key, offering limited variety can prevent monotony. BotVoice provides eight presets, allowing teams to assign different voices to different modules or languages without managing separate infrastructure. For example, use 'Aria' for English narration and 'Luna' for soft storytelling in other supported languages.
This approach allows for nuanced interactions, such as distinguishing between a system notification and a conversational agent.
- Assign voices by feature
- Distinguish agents from system sounds
Language Considerations
BotVoice supports 23 languages, allowing product teams to deploy voice assistants across diverse markets without managing separate voice assets. Any voice can speak any supported language, simplifying localization workflows for global chatbots and voice-first interfaces.
The API handles language selection through explicit codes or automatic detection. If the language field is omitted, the system guesses the language based on the text script. Latin script defaults to English, so non-English Latin text requires explicit specification to ensure accurate pronunciation and intonation.
Multilingual Support
The API supports 23 languages, including Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish.
All eight voice presets can generate audio in any of these languages, ensuring consistent tone and style across your multilingual product features.
- Supported: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
- Not supported: Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, Bengali, and others
Language Detection
When you do not specify a language, the API guesses it from the text script. Text in Latin script defaults to English if no language code is provided.
For non-English Latin-script text, always set the language parameter to ensure the correct voice is used and avoid mispronunciation of proper nouns or context-specific terms.
Localization
Use the language parameter in your JSON request body to enforce a specific locale. This ensures that your chatbot or voice assistant responds in the correct language for your target audience.
Combine this with the appropriate voice preset to maintain brand consistency, regardless of the user's region or preferred language.

Audio Format Choices
Choosing the right audio format balances file size, quality, and compatibility. Our API supports MP3, WAV, OPUS, FLAC, and PCM, letting you optimize for your specific delivery channel.
Select MP3 for broad device compatibility, WAV for uncompressed clarity, or OPUS for efficient bandwidth usage in voice assistants.
MP3 for Compatibility
MP3 is the universal standard for web and mobile applications. It offers a good balance between file size and audio fidelity, ensuring your voice content plays smoothly across all major platforms and devices without additional codecs.
WAV for Quality
WAV provides uncompressed, high-fidelity audio. Use this format when audio quality is critical, such as in high-end voice assistants or when further processing of the audio file is required without generation loss.
OPUS for Bandwidth
OPUS delivers excellent audio quality at lower bitrates. This format is ideal for voice-first interfaces where minimizing data usage and loading speed is essential, especially on mobile networks or in bandwidth-constrained environments.
Latency and Performance
Text-to-speech latency involves two distinct phases: the time required to synthesize audio from your input and the network delay in transmitting the resulting file. Understanding both helps product teams set realistic expectations for their voice assistants and chatbot interfaces.
Generation Time
Synthesis speed depends on text length and system load. Long texts process at roughly real-time speed (25 characters per second of audio), while short phrases may take a few seconds to generate. Requests are processed sequentially, so high demand can increase wait times.
Network Latency
After generation, the audio file is returned directly in the API response. Transmission time varies by region and connection quality but is generally minimal for standard audio formats like MP3 or WAV.
Queue Management
Under heavy load, requests may wait in a queue before processing begins. This is normal behavior for our cloud infrastructure and ensures fair resource distribution across all active keys.
- Rate limit: 60 requests per minute per key
- Initial limit: 10 requests per minute before first top-up
- No SLA or uptime guarantee
Data Privacy in TTS
BotVoice processes input text solely to generate the requested audio file. We do not store the generated audio in our systems after the response is delivered to your client. This ensures that your content remains transient and private during the synthesis process.
Input Data Usage
Your text is sent to the TTS engine, converted into audio, and returned in the response. No long-term storage of your specific text content occurs beyond what is necessary for immediate delivery.
No Training Data
The text you submit is used exclusively for synthesis. It is not used to train or improve our voice models. Your proprietary content remains yours, ensuring no data leakage into our training sets.
Audio Storage
Generated audio files are returned in the HTTP response and are not stored on our servers. Only request metadata (time, character count, status) is retained for billing and troubleshooting purposes.
Future of Voice Interfaces
Voice interfaces are shifting from novelty to standard utility. Product teams now prioritize voice as a primary input channel alongside text and touch, driven by the need for frictionless, hands-free user experiences in mobile and IoT applications.
This shift relies on reliable, low-latency TTS APIs that integrate seamlessly into existing chatbot and assistant architectures, ensuring consistent delivery across diverse devices and connection speeds.
Rise of Voice UX
Users increasingly expect natural, conversational responses from their devices. Integrating high-quality synthetic speech allows product teams to deliver richer, more engaging interactions without requiring complex audio engineering infrastructure.
Voice UX enhances accessibility and convenience, making applications usable in environments where visual attention is limited, such as during commutes or while cooking.
- Hands-free interaction for mobile and IoT
- Improved accessibility for visually impaired users
- Natural, conversational tone enhances engagement
AI Integration
Modern TTS APIs are designed to work alongside AI-driven intent recognition and dialogue management systems. This integration enables dynamic, context-aware responses that feel more human and responsive.
By decoupling voice synthesis from backend logic, developers can swap TTS providers or update voices without disrupting the core application logic or user data flows.
- Seamless integration with dialogue engines
- Dynamic response generation based on context
- Decoupled architecture for easier updates
Market Growth
The demand for voice-enabled applications is expanding as more users adopt voice assistants in homes and workplaces. Product teams are investing in voice features to meet this growing expectation for intuitive, spoken interactions.
As infrastructure costs drop and API reliability improves, more teams can afford to implement voice as a standard feature, driving further adoption across industries.
- Increasing adoption of voice assistants
- Growing demand for spoken interactions
- Reducing infrastructure costs for voice features
Common Misconceptions
Developers often assume TTS APIs provide a continuous audio stream, but the standard BotVoice endpoint returns a complete file. This distinction affects how you handle playback in voice assistants and chatbots, requiring you to manage file boundaries rather than listening to a live feed.
Another frequent confusion involves voice cloning and structured markup. Our service uses eight fixed presets and does not offer custom voice cloning or SSML support, so you must rely on standard text inputs and speed adjustments rather than complex tags or cloned audio profiles.
Streaming vs File Download
The API returns a finished audio file (mp3, wav, opus, etc.) rather than a WebSocket stream. You receive the full asset after synthesis completes, which simplifies integration but means you cannot listen to partial audio in real-time.
Custom Voice Cloning
We do not offer voice cloning. You select from eight licensed presets (e.g., Aria, Felix, Luna) based on their existing tones. There is no option to upload your own voice sample to create a unique digital twin.
SSML Support
The endpoint accepts plain text input, not SSML markup. You cannot use tags for pauses, emphasis, or pronunciation rules. Control is limited to the speed parameter (0.25–4.0) and voice selection.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Build conversational flows that feel human to delight users and drive engagement.
23 languages
Any voice speaks any of these languages.
