Frequently asked questions
What is the character limit for a single request?
Each API request accepts a maximum of 4096 characters in the input field. This limit applies to the total text length, including spaces and punctuation. If your content exceeds this threshold, you must split the text into smaller parts and send multiple requests to ensure successful synthesis.
Can I use the OpenAI SDK with HomeEcho?
Yes, the official OpenAI SDKs work seamlessly with our API. You only need to change the base_url to https://api.playbk3.com/v1 and use your HomeEcho API key. The request shape matches the OpenAI speech endpoint, allowing you to integrate our TTS engine into existing workflows with minimal code changes.
Which voices support which languages?
All eight voice presets support all 23 supported languages. You can use any voice, such as Aria or Felix, to speak Arabic, Chinese, English, or any other supported language. If the language is not specified, the system guesses it, but you should set it explicitly for non-English Latin text.
Is the API suitable for real-time voice assistants?
The API does not support streaming or WebSocket connections. Requests are processed sequentially, and generation takes roughly real-time speed (25 characters per second). For voice assistants, you should pre-generate audio files or handle short phrases with acceptable latency, as requests may wait in a queue under load.
How do I handle texts longer than 4096 characters?
You must split the input text into parts of up to 4096 characters each. Send each part as a separate POST request to the speech endpoint. Then, join the resulting audio files on your side to create the final continuous audio output for your IoT device or application.
What audio formats are supported?
The API supports five audio formats: mp3, wav, opus, flac, and pcm. You can specify your preferred format in the response_format field of your request. This allows you to choose the best balance between file size and quality for your specific IoT device or network constraints.
Does the API support SSML tags?
No, the API does not support SSML. It accepts plain text input only. If you need specific pronunciation or pacing, you must rely on the voice presets and speed adjustments, as there are no phoneme dictionaries or custom lexicons available for fine-tuning.
How is pricing calculated for IoT devices?
Pricing is based on character count at $10 per 1 million characters, or $0.01 per 1,000 characters. Spaces and punctuation count toward the total. Failed requests are free, so you only pay for successful synthesis. This model allows predictable cost management for bulk audio generation in your IoT deployments.
Can I clone my own voice for device notifications?
No, we do not offer custom voice cloning. You cannot upload your own voice or design a new voice. You must choose from the eight available presets (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion) which are based on licensed recordings from LibriTTS-R and LibriVox.
Is there a free trial available?
Yes, signing in with Google provides a $0.50 trial credit, valid for 7 days, which covers approximately 50,000 characters. Accounts created with email and password do not receive this trial credit. You can test the API thoroughly before committing to a top-up for your production IoT environment.
What payment methods are accepted?
We accept USDT on the TRON network (TRC-20) for top-ups. There are no credit cards, PayPal, or subscriptions. The minimum top-up is $5, and funds are credited automatically within minutes. Balance does not expire, and there are no monthly plans or invoices with VAT for this payment method.
Do you offer an SLA for uptime?
No, we do not offer an SLA or uptime guarantee. There is no enterprise contract or on-premise install option. Support is by email only. We recommend implementing retry logic in your IoT devices to handle potential latency or queue delays.
How long are request logs kept?
Request metadata, including time, character count, and status, is kept for billing purposes. Input text is used only to produce the audio and is not stored for training. Generated audio is returned in the response and not stored on our servers, ensuring privacy for your device notification data.
Can I generate audio in bulk for firmware updates?
Yes, you can generate audio in bulk by splitting long texts and sending multiple requests. With a rate limit of 60 requests per minute, you can efficiently process large volumes of text. Failed requests are free, ensuring you only pay for successful audio files generated for your firmware update notifications.
First request
curl
curl https://api.playbk3.com/v1/audio/speech \
-H "Authorization: Bearer $TTS_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"tts-1","voice":"aria","input":"Hello from the API.","response_format":"mp3"}' \
--output hello.mp3Why IoT Developers Choose HomeEcho
HomeEcho delivers reliable text-to-speech synthesis for smart home and IoT developers who require consistent audio output without external dependencies. Our API handles voice generation at scale, ensuring deterministic results for voice assistants and device notifications.
Deterministic Audio Generation
Each synthesis request produces a consistent audio file based on the input text, voice preset, and speed parameters. This predictability ensures that voice commands and device alerts sound uniform across different hardware units and firmware versions.
Low Latency for Voice Assistants
Text is processed sequentially, with typical speed generating roughly 25 characters of text per second of audio. Short phrases complete in a few seconds, while longer texts approach real-time generation, minimizing wait times for interactive voice responses.
Standard REST API Integration
Integration requires a single POST request to the /v1/audio/speech endpoint. The API accepts JSON bodies identical to the OpenAI speech endpoint shape, allowing developers to reuse existing HTTP clients or SDKs by simply updating the base URL.
Voice Engine Specifications
The HomeEcho engine delivers consistent, deterministic audio synthesis optimized for smart home devices and IoT voice assistants. By using fixed presets rather than variable generative models, we ensure that voice output remains identical across repeated requests, which is critical for reliable device notifications and standardized user experiences.
Our architecture supports 23 languages and multiple audio formats, allowing developers to deploy globally without managing complex localization pipelines. The system processes requests sequentially, ensuring stable throughput for high-volume IoT deployments.
All generated audio files carry an inaudible watermark to clearly identify synthetic speech, maintaining transparency for end-users. This deterministic approach eliminates the variability often found in cloud-based TTS services, providing the consistency required for professional-grade voice interactions.
8 Professional Presets
HomeEcho provides eight voice presets trained on licensed LibriTTS-R recordings and public-domain LibriVox readings. These voices are engineered for clarity and distinct tonal characteristics, ensuring that voice assistants sound professional and recognizable.
Each voice is optimized for specific use cases, from narrative storytelling to energetic promotional reads. The deterministic nature of these presets guarantees that a specific voice ID always produces the same acoustic output for the same input text.
- Female: Aria (warm narrator), Nova (bright upbeat), Luna (soft storytelling), Sage (calm instructional)
- Male: Felix (energetic promo), Milo (friendly conversational), Atlas (neutral news), Orion (deep documentary)
23 Supported Languages
The API supports 23 languages, covering major global markets including English, Chinese, Spanish, German, Japanese, and Hindi. Any voice can speak any of these languages, allowing developers to mix and match voices and languages without additional configuration overhead.
For Latin-script texts outside of English, the language parameter should be set explicitly to ensure accurate pronunciation. The system guesses the language from the script if not provided, defaulting to English for Latin text.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
Multiple Audio Formats
Developers can choose from five response formats: MP3, WAV, Opus, FLAC, and PCM. MP3 is recommended for web delivery and standard IoT device playback due to its balance of quality and file size. WAV provides uncompressed local playback options.
Opus and FLAC offer high-quality compression for bandwidth-constrained environments, while PCM provides raw audio data for immediate processing. The API returns the audio file directly in the response, eliminating the need for secondary download steps.
- MP3 for web delivery and standard IoT playback
- WAV for uncompressed local playback
- Opus and FLAC for high-quality compression
- PCM for raw audio data processing
Technical Implementation
Integrate HomeEcho’s TTS API into your IoT stack using standard RESTful principles. The endpoint mirrors the OpenAI speech API structure, allowing you to leverage existing OpenAI SDKs by simply adjusting the base URL to https://api.playbk3.com/v1. This compatibility reduces development time for teams already familiar with modern API consumption patterns.
OpenAI SDK Compatibility
Use the official OpenAI SDKs with a custom base URL configuration. Send requests to https://api.playbk3.com/v1/audio/speech and receive audio files directly in your application logic. This approach eliminates the need for custom HTTP clients or complex parsing libraries, ensuring consistent behavior across different programming environments.
- Compatible with standard OpenAI SDKs
- Simple base URL modification
- Standard JSON request/response format
JSON Request Body
Submit text via a JSON body containing the model, input text, voice selection, and optional parameters like speed or language. The API accepts up to 4096 characters per request. For longer texts, split the input and join the resulting audio files to ensure deterministic output for voice assistants.
- Model: tts-1 or tts-1-hd
- Input text up to 4096 characters
- Optional speed (0.25–4.0) and language
Bearer Token Authentication
Secure your API calls using a Bearer token in the Authorization header. Each account holds one active key; issuing a new key immediately invalidates the previous one. This single-key model simplifies credential management for distributed IoT devices and backend services without requiring complex OAuth flows.
- Header: Authorization: Bearer tts-...
- One active key per account
- Instant invalidation on key refresh
Supported Audio Formats
HomeEcho provides four distinct audio formats to match your IoT device’s playback capabilities. Selecting the correct format ensures optimal compatibility between your smart home infrastructure and the generated voice output.
MP3 for Web Delivery
MP3 is the standard format for web-based delivery and most modern IoT devices. It offers a balance between file size and quality, making it suitable for voice assistants that play audio over HTTP.
WAV for Local Playback
WAV provides uncompressed audio for local playback on devices that require high-fidelity, lossless data. This format is suitable for stationary smart speakers or hubs that process audio locally without network constraints.
FLAC and Opus for Quality
FLAC and Opus offer lossless and efficient lossy compression respectively. Use FLAC for archival quality or Opus for bandwidth-constrained environments where clear speech synthesis is critical without excessive data usage.
- FLAC: Lossless compression for archival or high-end playback
- Opus: Efficient compression for low-bandwidth IoT connections
- MP3: Standard for web streaming and general use
- WAV: Uncompressed for local, high-fidelity playback
Pricing Structure
HomeEcho utilizes a transparent, per-character billing model that aligns costs directly with usage. The standard rate is $10 per 1 million characters, equating to $0.01 per 1,000 characters. Spaces and punctuation are counted as characters, ensuring predictable expenses for smart device notifications and voice assistant outputs.
Per-Character Billing
Costs are calculated based on the exact character count of the input text. Failed requests incur no charge, and credits are returned in full if synthesis fails. This structure allows developers to budget accurately for high-volume IoT deployments without worrying about hidden per-minute or per-request fees.
- Rate: $10 per 1,000,000 characters ($0.01 per 1,000 chars)
- Spaces and punctuation count toward the total
- Failed requests are free with full credit return
No Monthly Fees
There are no monthly subscriptions, seats, or minimum spend requirements. Accounts can remain inactive without incurring charges. This pay-as-you-go model is ideal for indie developers and IoT projects with variable traffic patterns, ensuring you only pay for the audio generated by your devices.
USDT Payment Only
Top-ups are processed exclusively via USDT on the TRON network (TRC-20). The minimum top-up amount is $5, and credits are credited automatically within minutes of transfer confirmation. We do not support credit cards, PayPal, or invoices with VAT, maintaining a streamlined, borderless payment flow for global IoT deployments.
- Accepted: USDT (TRC-20) on TRON network
- Minimum top-up: $5
- Credits auto-credit within minutes; balance never expires

Voice Options Overview
HomeEcho offers eight distinct voice presets engineered for consistent audio output in smart home environments. These voices are derived from licensed recordings, ensuring high-quality synthesis for device notifications and voice assistants without the need for external dependencies.
Female Voices: Aria, Nova
Aria provides a warm, clear narrator tone suitable for instructional content. Nova delivers a bright, upbeat delivery ideal for active device interactions.
Male Voices: Felix, Milo
Felix offers an energetic, promo-style voice for dynamic alerts. Milo provides a friendly, conversational tone for general-purpose notifications.
Neutral and Deep Voices
Atlas ensures neutral, news-style clarity for objective information delivery. Orion provides a deep, documentary-style voice for detailed audio content.
Language Support
HomeEcho supports 23 languages, covering major Latin, Cyrillic, and Asian scripts. Developers can generate speech in Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. Each voice handles all supported languages without requiring specific model variants.
Latin Script Languages
Standard Latin-based languages are fully supported across all eight voice presets. This includes English, French, German, Spanish, Portuguese, Italian, Dutch, Swedish, Norwegian, Danish, Finnish, and Polish. The engine handles standard character encoding without requiring special configuration for Western European markets.
Cyrillic and Asian Scripts
Non-Latin scripts are processed with equal precision. Supported Cyrillic languages include Russian, while Asian script support covers Chinese, Japanese, and Korean. Additional coverage includes Arabic, Hebrew, Hindi, Greek, Malay, Swahili, and Turkish, ensuring broad geographic deployment for IoT devices.
Automatic Language Detection
The API guesses the language from the input script if the language field is omitted. Latin script defaults to English. For non-English text using Latin characters, explicitly set the language parameter to ensure accurate pronunciation and prevent misinterpretation.
Developer Experience
HomeEcho provides a straightforward integration path for IoT applications, allowing developers to generate speech with minimal code changes. The API follows standard REST conventions, ensuring that existing HTTP clients and libraries can interact with the service without specialized drivers.
Quick Start Guide
Begin by obtaining an API key via the /get-api-key/ page. Use the standard OpenAI SDK or any HTTP client to send a POST request to the /v1/audio/speech endpoint. Pass your text in the input field and select a voice preset from the 8 available options.
cURL Examples
The request structure mirrors standard JSON bodies, making it easy to adapt from existing examples. You can test the endpoint immediately using the command line, passing your Bearer token in the Authorization header and specifying the desired response format.
Rate Limit Transparency
New keys start with a limit of 10 requests per minute before the first top-up. After the first top-up, this increases to 60 requests per minute. This predictable throttling helps IoT devices manage bandwidth and processing loads without unexpected failures.
Data and Privacy
HomeEcho processes input text exclusively for immediate audio synthesis. We do not use your input data to train our voice models, ensuring that your proprietary content remains isolated from our training datasets. This guarantees consistency and prevents your unique voice data from influencing future generations.
No Training on Input Text
Your text is used solely to generate the requested audio file. Once synthesis is complete, the input data is not retained for model improvement or secondary analysis. This separation ensures that your device notifications or voice assistant scripts remain distinct from our core engine data.
- Input text is not used for training
- Data remains isolated from model updates
- No secondary analysis of input content
Transient Audio Storage
Generated audio files are returned directly in the API response and are not stored on our servers. We retain only request metadata—such as timestamp, character count, and status—for billing purposes. This ensures minimal data footprint and simplifies compliance for IoT deployments.
Synthetic Watermarking
Every generated audio file includes an inaudible digital watermark. This mark identifies the output as synthetic speech, providing a clear distinction between machine-generated audio and human recordings. This feature aids in content verification for your smart home and IoT applications.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence.
23 languages
Any voice speaks any of these languages.
