Frequently asked questions
Which voice is best for a smart thermostat?
Atlas suits neutral status updates, while Sage works well for calm, instructional guidance. Both voices support all 23 languages, ensuring consistent tone for temperature changes. Select the voice that matches your device’s personality, keeping interactions simple and unobtrusive for daily adjustments. There is no single "best" voice; choose based on your specific interface needs.
How do I choose between Aria and Luna?
Choose Aria for a warm, clear narrator style, ideal for straightforward instructions. Pick Luna for a soft, storytelling tone, perfect for ambient notifications or evening routines. Both voices are available in all supported languages, allowing you to match the mood of your smart home environment.
Is Felix suitable for a security camera?
Felix offers an energetic, promo-style voice suitable for alert notifications. Use him for motion alerts or package deliveries where immediate attention is needed. His tone conveys urgency without being alarming, fitting well with modern IoT security interfaces that require clear, active communication.
What language should I use for global devices?
Set the language parameter to the specific ISO code for each device, as Latin text defaults to English. Any voice can speak any of the 23 supported languages. Explicitly defining the language ensures accurate pronunciation and prevents unintended English defaults for non-English text.
How does speed affect user experience?
Adjust speed between 0.25 and 4.0 to control pacing. Faster speeds suit quick alerts, while slower rates improve clarity for complex instructions. Since requests are processed sequentially, higher speeds may reduce generation time, but queue delays can occur under load, affecting real-time responsiveness.
Can I use multiple voices in one app?
Yes, you can assign different voices to different events within your application. Each API call accepts a specific voice parameter, allowing you to mix Aria for greetings and Felix for alerts. This flexibility lets you create distinct audio identities for various smart home functions.
What is the best format for mobile apps?
MP3 is suitable for mobile apps due to its balance of quality and file size. FLAC offers lossless quality, while OPUS provides efficient compression. PCM is uncompressed but larger. Choose based on your bandwidth constraints and quality requirements, ensuring smooth playback on IoT devices.
How do I handle long notifications?
Split text into parts of up to 4096 characters per request. Generate audio for each part and join the files in your application. This approach ensures no data loss and maintains audio continuity, allowing you to deliver lengthy instructions or stories without API limits.
Is the API reliable for live devices?
The API processes requests sequentially, so under load, requests may wait in a queue. Typical speed is roughly real time for long texts, but short phrases take a few seconds. There is no SLA or uptime guarantee, so design your app to handle potential delays gracefully.
How much does it cost to equip 100 devices?
Cost depends on usage, not devices. You pay $10 per 1 million characters. If each device generates 1000 characters daily, 100 devices use 100,000 characters daily, costing $1.00 per day. Balance does not expire, and failed requests are free, so costs scale with actual generation.
Do voices sound natural in Spanish?
Yes, all 8 voices support Spanish, one of the 23 supported languages. Voices are based on licensed recordings and carry an inaudible synthetic watermark. Set the language parameter to 'es' to ensure correct pronunciation, as Latin text defaults to English if not specified.
Can I adjust volume via API?
No, the API does not support volume adjustment. It returns the audio file in the requested format (mp3, wav, opus, flac, pcm). You must handle volume control within your application or device hardware after receiving the generated audio file from the endpoint.
What happens if the internet cuts out?
If the internet cuts out during generation, the request may fail or time out. Failed requests are free, so you are not charged for incomplete audio. Your application should handle retries or cache the last successful audio file to ensure continuous operation during connectivity issues.
Do I need to update voices manually?
No, voices are static presets (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion). They do not change automatically. You select the voice in each API request. To change a voice, update your application’s configuration to call the API with the desired voice parameter.
Understand Your User
Smart home users range from tech-savvy early adopters to elderly residents needing clear, slow-paced audio. Your voice engine must adapt to these varying needs without requiring complex configuration. Understanding who will hear your device’s prompts ensures the tone matches their comfort level, whether it’s a quick notification for a busy professional or a detailed instruction for a senior user.
Consider the environment where the device operates. Kitchen noise, bathroom echoes, or quiet bedrooms all affect how speech is perceived. A warm, clear narrator like Aria works well for general instructions, while a calm voice like Sage suits bedtime routines. Matching the voice to the physical context reduces cognitive load and improves user acceptance.
Demographics
Target your primary user group’s age and tech literacy. Younger users often prefer energetic, fast-paced voices like Felix for quick updates. Older demographics may find Sage’s calm, instructional tone easier to process. Avoid overly dramatic tones that might feel unnatural in a home setting.
- Prioritize clarity over speed for elderly users.
- Use warm tones for general home assistants.
- Keep notifications brief and friendly.
Context of Use
Sound environments vary significantly. In a noisy living room, a bright voice like Nova cuts through background noise better than a soft whisper. In a bedroom, Luna’s soft storytelling style feels less jarring. Adjust your voice selection based on whether the device is primarily used for entertainment, information, or control.
Remember that audio clarity is key. Even with noise-canceling features, the voice itself should remain distinct. Test your chosen voice in actual conditions to ensure it doesn’t sound robotic or overly synthetic in specific acoustic environments.
Voice Presets Overview
Selecting the right voice defines your device's personality. Our eight presets cover distinct tones, from warm narration to energetic prompts, ensuring your smart home interface feels natural and consistent.
Each voice is engineered for clarity in domestic environments. Choose based on whether your app requires a calm instructional tone or a bright, upbeat alert style for everyday interactions.
Female Voices
Aria offers a warm, clear narrator style ideal for step-by-step instructions. Nova provides a bright, upbeat tone suitable for quick status updates. Luna’s soft delivery works well for storytelling modes, while Sage ensures calm, instructional clarity for complex settings.
- Aria: Warm, clear narrator
- Nova: Bright, upbeat
- Luna: Soft, storytelling
- Sage: Calm, instructional
Male Voices
Felix delivers an energetic, promo-style voice perfect for promotional alerts or active mode announcements. Milo offers a friendly, conversational tone that feels like a helpful assistant. Atlas remains neutral and news-like for factual updates, while Orion provides a deep, documentary style for immersive experiences.
- Felix: Energetic, promo
- Milo: Friendly, conversational
- Atlas: Neutral, news
- Orion: Deep, documentary
Language Support
Our API supports 23 languages, ensuring your smart home device can communicate clearly with users worldwide. Any voice preset can speak any of these languages, allowing you to maintain a consistent brand personality across different regions without managing multiple voice libraries.
This flexibility simplifies development for indie developers targeting global markets. You can deploy the same voice engine for a thermostat in Berlin and one in Tokyo, ensuring a cohesive user experience regardless of the local language.
23 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.
If the language field is omitted, the system guesses the language from the script. Latin text defaults to English, so explicitly set the language for other Latin-script texts to avoid mispronunciation.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
Spanish Specifics
Spanish is fully supported with all 8 voice presets. Whether you need the warm narration of Aria or the calm instruction of Sage, Spanish users will receive clear, natural speech.
Ensure you set the language parameter to es in your API requests. This prevents the system from defaulting to English pronunciation rules, which can cause awkward accents in Spanish text.
imageCaption
Audio Formats
Choosing the right audio format balances clarity, file size, and device storage. MP3 offers efficient compression for network transmission, while WAV provides lossless quality for local playback. Your choice depends on whether you prioritize bandwidth savings or maximum fidelity for your smart home interface.
Understanding file size impact helps you optimize storage on constrained IoT devices. Larger files ensure higher quality but consume more memory and bandwidth. Selecting the appropriate format ensures smooth voice responses without draining device resources or network limits.
MP3 vs WAV
MP3 is ideal for network transmission over Wi-Fi or cellular networks. It reduces data usage significantly, making it suitable for devices with limited bandwidth. WAV provides lossless quality for local playback.
WAV files are larger but offer pristine sound quality. They are best used when audio is played locally on the device or when network conditions are stable. MP3 is efficient for network transmission.
- MP3: Compressed, smaller size, ideal for network streaming
- WAV: Uncompressed, larger size, best for local high-fidelity playback
- OPUS: Efficient compression, good for low-bandwidth environments
- FLAC: Lossless compression, balances quality and size
File Size Impact
File size directly affects storage requirements on your IoT device. Larger audio files may limit the number of voice prompts you can store locally. Smaller files allow for more extensive voice libraries without requiring frequent updates or cloud dependencies.
Consider the trade-off between audio quality and storage constraints. For devices with limited memory, MP3 is often the optimal choice. For devices with ample storage, WAV ensures the best possible listening experience for your users.
- Larger files consume more device storage
- Smaller files reduce network bandwidth usage
- Balance quality with storage limitations
- Optimize for frequent updates vs. static content
Technical Constraints
Building reliable smart home interfaces requires understanding the technical boundaries of our TTS API. Developers must account for request throughput and payload size to ensure smooth voice interactions without hitting system limits during peak usage.
Rate Limits
Our API allows 60 requests per minute per active key, which scales well for most IoT deployments. New accounts start with a lower limit of 10 requests per minute until the first top-up is processed.
- Standard limit: 60 requests/minute per key
- Initial limit: 10 requests/minute for new accounts
- Limit resets every 60 seconds
Character Limits
Each request supports up to 4096 characters of input text. For longer prompts, split the text into smaller chunks and join the resulting audio files on the client side. This ensures consistent latency and prevents timeout errors.
Pricing Strategies
Understanding the per-device cost helps you budget for hardware deployments. With our API, pricing is straightforward: $0.01 per 1,000 characters. Since smart home devices often send short commands, costs remain low per interaction. Failed requests incur no charge, protecting your margins during testing phases.
Volume discounts are not offered as a separate tier. Instead, your balance scales naturally with usage. Top-ups are processed instantly via USDT (TRC-20), ensuring you never run out of credit mid-deployment. No monthly plans or minimum spends mean you only pay for what your users actually hear.
Cost per Device
Calculate your average request length to estimate hardware expenses. A typical thermostat command might use 50 characters, costing less than $0.001 per use. This predictability allows precise pricing models for your IoT service.
Include punctuation and spaces in your character counts. Each character contributes to the total, so optimizing command phrasing can slightly reduce long-term costs without sacrificing clarity.
- $0.01 per 1,000 characters
- Failed requests are free
- Balance does not expire
Volume Discounts
Our model eliminates complex tiered pricing. You simply top up your account when needed. The minimum top-up is $5, making it easy to start small and scale as your device fleet grows.
No subscriptions or seats mean you retain full control over your spending. Use the dashboard to monitor character consumption and adjust your budget strategy based on real usage data.
Integration Steps
Integrating PPT Gaming Voices into your IoT device is straightforward. You begin by generating an API key on the /get-api-key/ page. This key authenticates your requests to the TTS engine, ensuring only your devices consume your character balance. Keep this key secure, as a new issuance immediately invalidates the old one.
Once you have your credentials, configure your development environment. The API follows the standard OpenAI request shape, allowing you to use existing OpenAI SDKs by simply pointing them to our base URL: https://api.pptgaming.info/v1. This minimizes code changes and leverages familiar tooling for your smart home app.
API Key Setup
Navigate to /get-api-key/ and sign in with Google or email. Your unique Bearer token appears immediately. Use this token in the Authorization header for every POST request to /v1/audio/speech. Remember, you only need one active key per account, simplifying device management.
- Sign up via /get-api-key/ using Google or email/password
- Retrieve your Bearer token from the dashboard
- Note: issuing a new key instantly revokes the old one
SDK Configuration
Use the official OpenAI SDKs by changing the base_url to https://api.pptgaming.info/v1. Pass your text, selected voice (e.g., 'aria' or 'atlas'), and desired format. The API returns the audio file directly, ready for playback on your device.
Ensure you handle the 60 requests-per-minute rate limit. For devices generating many short alerts, queue your requests to avoid throttling. The API returns standard HTTP status codes to help you manage these constraints effectively.
- Set base_url to https://api.pptgaming.info/v1
- Use standard OpenAI SDKs with your Bearer token
- Respect the 60 requests/minute limit per key
Testing and QA
Testing ensures your smart home interface sounds natural and responsive. Use the /v1/voices endpoint to preview the eight available presets, selecting styles like Luna for storytelling or Felix for quick alerts. Verify that the voice matches your device’s personality before committing to production.
Voice Samples
Generate short audio clips using the API with your target text. Check that the tone remains consistent across different languages, such as English and Spanish, since any voice can speak any supported language. This step helps you avoid mismatched emotional cues in your final product.
Latency Checks
Measure the time from request to audio delivery. Short phrases may take a few seconds, while longer texts generate at roughly real-time speed. Monitor queue delays during peak hours to ensure your device provides timely feedback without frustrating the user.
- Test with 4096-character limits to verify maximum load handling.
- Monitor response times for different response formats like MP3 or WAV.
Deployment
Deploying your TTS engine to smart home devices requires careful consideration of latency and update mechanisms. Since requests are processed sequentially, ensure your device handles queueing gracefully during peak usage. Use the provided curl examples to validate your integration before mass deployment.
Live Updates
You can switch voices instantly via API parameters without redeploying code. This allows you to A/B test different tones across your user base easily.
- Test new voice presets in staging before rolling out to production devices.
- Update voice parameters via API calls without requiring a firmware patch.
Monitoring
Track character usage and request status via your dashboard. Since failed requests are free, monitor error rates to identify specific voice or format issues. Keep an eye on rate limits (60 requests/minute) to prevent throttling during high-traffic events.
Use the request metadata logs to audit billing and diagnose latency issues. Ensure your devices handle 429 Too Many Requests errors gracefully by implementing exponential backoff.
- Monitor dashboard for character consumption and balance.
- Implement exponential backoff for rate limit errors.
Common Use Cases
Smart home interfaces demand clarity above all. Your thermostat or security camera isn’t a podcast; it’s a utility. Users need immediate, unambiguous instructions without cognitive load. This means choosing voices that prioritize articulation over emotional warmth, ensuring commands are understood instantly even in noisy environments.
For devices operating 24/7, consistency matters. A voice that remains neutral and steady prevents user fatigue during repeated interactions, like nightly arming sequences or temperature adjustments. The goal is seamless integration where the voice feels like part of the hardware, not a separate app layer.
Thermostats
Thermostats require precise, calm instruction. Use Atlas for neutral, news-style clarity during status updates, or Sage for gentle, instructional guidance when adjusting settings. These voices ensure users understand temperature changes without distraction.
Avoid high-energy voices like Felix for routine adjustments. A calm tone reduces cognitive load, making daily interactions feel natural rather than robotic. This approach supports long-term user comfort.
- Use Atlas for clear status updates
- Choose Sage for gentle instructional guidance
- Avoid energetic tones for routine adjustments
Security Cameras
Security cameras need immediate attention. Orion’s deep, documentary tone conveys authority during alerts, while Milo’s friendly conversational style works well for two-way talk. Both ensure critical messages are heard clearly.
For motion alerts, clarity is key. Use Orion to signal importance without causing alarm. For two-way communication, Milo’s approachable tone encourages user engagement, making interactions feel personal yet secure.
- Use Orion for authoritative alerts
- Choose Milo for friendly two-way talk
- Prioritize clarity over emotional warmth
Troubleshooting
When your smart device speaks the wrong language, it is usually because the API guessed incorrectly from Latin script. If your text is in Dutch or Swedish but lacks a language tag, the system defaults to English. Always set the language field explicitly to avoid this confusion for non-English Latin texts.
Audio glitches often stem from network timeouts or exceeding character limits. Since requests are processed sequentially, high load can cause delays. Ensure your client handles partial responses gracefully and splits long texts into 4096-character chunks to maintain consistent playback quality.
Language Errors
Ensure the language parameter matches your input text. The API guesses from script, so Latin text defaults to English unless specified.
Unsupported languages like Vietnamese or Thai are not supported. Stick to the 23 supported languages listed in our docs.
- Set
languageexplicitly for non-English Latin scripts. - Check supported list: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish.
Audio Glitches
High server load may delay synthesis. Requests queue sequentially, so long texts take longer. Split inputs into 4096-character chunks and join files.
Failed requests return zero characters and are free. Check your status code before playing audio to avoid silent failures.
- Split long texts into 4096-character parts.
- Handle queue delays with exponential backoff.
- Verify
statusbefore playback.
Future Proofing
Plan for growth by leveraging our unified API architecture. Since the same endpoint handles all voices and languages, adding new capabilities requires zero code changes. Your infrastructure scales seamlessly as your device fleet expands, ensuring consistent performance without complex migrations.
Stay current with language support without waiting for major releases. The API currently supports 23 languages, allowing you to deploy globally from day one. New language additions are rolled out to the existing endpoint, keeping your integration stable and future-ready.
Scalability
Design for peak loads by respecting the 60 requests-per-minute limit per key. For large-scale deployments, issue multiple API keys to parallelize synthesis. This approach maintains low latency even when thousands of devices request audio simultaneously, ensuring smooth user experiences.
- Use multiple keys to bypass per-key rate limits
- Batch split long texts to manage queue times
- Monitor usage dashboards to optimize key distribution
New Languages
Leverage the broad language support covering Arabic, Chinese, French, German, Hindi, Japanese, Korean, Spanish, and more. Since any voice can speak any supported language, you can easily localize your smart home interface for new markets without updating your core application logic.
- Support for 23 distinct languages
- Automatic language guessing for Latin scripts
- Explicit language codes ensure accuracy for non-English Latin text
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Even with noise-canceling features, the voice itself should remain distinct.
23 languages
Any voice speaks any of these languages.
