Frequently asked questions
What is text to speech?
Text-to-speech (TTS) is a cloud API that converts written text into audio files. You send a JSON request with your text and voice preference, and receive an audio file in return. It is designed for scalable audio content generation without recording human voice actors for every language or variation.
How does text to speech work for museum audioguides?
Send your exhibit descriptions via POST to /v1/audio/speech. Specify the desired voice and language. The API returns an audio file you can embed in your museum app or website. This allows you to produce high-quality audioguides in multiple languages without hiring voice actors for each translation.
Which voice is best for a museum tour guide?
For museum guides, atlas (neutral, news) or sage (calm, instructional) are ideal. Atlas provides a clear, neutral tone suitable for factual exhibits. Sage offers a calm, instructional style perfect for detailed explanations. Both voices are available in all supported languages, ensuring consistency across your collection.
Can I use this API for hotel concierge audio messages?
Yes. You can generate concise audio messages for hotel services like room service or check-in instructions. Send text up to 4096 characters per request. The API returns high-quality audio files that can be integrated into your hotel’s digital concierge system or mobile app for guest assistance.
How many languages does the TTS API support?
The API supports 23 languages, including English, Spanish, French, German, Japanese, and Chinese. Any voice can speak any of these languages. If you do not specify a language, the system guesses it from the script, defaulting to English for Latin text. Vietnamese and Thai are not supported.
Is the text used to train the voice model?
No. Input text is used only to produce the specific audio file returned in the response. It is not used for training the underlying voice model. Generated audio is not stored by the API. Only request metadata, such as time and character count, is kept for billing purposes.
What happens if the API request fails?
If synthesis fails, the charge is returned in full. You are not billed for failed requests. This ensures you only pay for successfully generated audio files. You can retry the request or adjust parameters like speed or voice to resolve the issue without financial penalty.
Do I need to pay monthly for the TTS service?
No. There are no monthly plans, subscriptions, or minimum spends. You pay $10 per 1 million characters. You top up your balance using USDT on the TRON network. The balance does not expire, and you only pay for the characters you actually generate.
Can I get a custom voice for my brand?
No. The API offers 8 preset voices based on licensed recordings from LibriTTS-R and LibriVox. There is no custom voice cloning or uploading of your own voice. You cannot design a new voice or modify existing ones. You must choose from the available male and female presets.
How long does it take to generate an audio file?
Typical speed is roughly real time, about 25 characters per second of generation. Short phrases may take a few seconds. Requests are processed sequentially, so under load, a request may wait in a queue. Longer texts are split into parts of up to 4096 characters and joined by the client.
What audio formats are supported for tourism videos?
The API supports mp3, wav, opus, flac, and pcm formats. You can specify the format in your request. MP3 and WAV are widely compatible with most tourism video players and apps. Choose the format that best fits your distribution channel and quality requirements.
Is there a rate limit for generating audioguides?
Yes, the limit is 60 requests per minute per API key. Before your first top-up, the limit is 10 requests per minute. If you generate many audioguides simultaneously, requests may queue. Plan your generation workflow to stay within these limits for consistent performance.
Can I use this for YouTube voiceovers?
Yes. You can generate voiceovers for YouTube videos using any of the 8 voices. Each generated file carries an inaudible watermark marking it as synthetic speech. You pay only for the characters used in your scripts.
How do I handle texts longer than 4096 characters?
Split your text into parts of up to 4096 characters. Send each part as a separate request to the API. Then, join the resulting audio files in your application. This ensures each request stays within the character limit while producing a complete audio file.
First request
curl
curl https://api.a2229.info/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 TTS Matters for Tourism
Tourism startups and museums face a recurring challenge: producing high-quality audio content for dozens of locations and languages without breaking the budget. Traditional voice actor sessions require significant upfront investment and scheduling coordination, often delaying product launches for months.
Scaling Audio Guides Without Voice Actors
Generating audio via API removes the need to book studios for every new exhibit or language variant. You can produce hundreds of minutes of content in hours, scaling your tour offerings instantly based on visitor demand.
- Eliminate studio booking costs
- Produce content on demand
- Zero upfront voice actor fees
Consistent Brand Voice Across Locations
Maintaining a uniform tone across multiple branches is difficult with human actors who may interpret scripts differently. Our API ensures that every visitor, whether in Tokyo or Berlin, hears the exact same narration style, reinforcing brand identity.
Reducing Production Time for Video Tours
Video tours often require voiceovers that are time-consuming to record and edit. With a simple REST API call, you can generate audio files in parallel, drastically cutting the time from script to final media asset.
How Our API Works for Museum Audio
Tourism startups and museum curators can integrate audio generation directly into their content workflows using a standard REST API. By sending a POST request with your script, you receive an audio file ready for deployment in apps, websites, or digital signage.
This approach eliminates the logistical overhead of booking voice talent for every new language or update. You generate content on demand, scaling your tour offerings based on visitor demand without waiting for external production cycles.
Simple REST API Integration
Use the standard OpenAI-compatible endpoint at https://api.a2229.info/v1/audio/speech. Send a JSON body with model, input, and voice. The API returns the audio file directly, allowing you to store it or stream it to your users.
With a simple REST API call, you can generate audio files, cutting the time from script to final media asset. Requests are processed sequentially, so queue times may apply during high load, but results are reliable.
- Endpoint:
POST /v1/audio/speech - Format: JSON request, audio response
- SDK: Works with official OpenAI SDKs by changing
base_url
8 Professional Narrator Voices
Select from eight distinct presets ranging from warm narrators like Aria and Luna to neutral voices like Atlas and Orion. Each voice is based on licensed recordings from LibriTTS-R and LibriVox, ensuring high-quality, consistent narration styles suitable for audioguides.
Voices like Milo offer a friendly, conversational tone for interactive tours, while Sage provides calm, instructional clarity for complex exhibits. All voices support the full range of 23 languages, allowing you to maintain a consistent brand voice across locations.
- Aria: Warm, clear narrator
- Atlas: Neutral, news-style
- Milo: Friendly, conversational
- Luna: Soft, storytelling
Multi-Language Support for Global Tours
Support 23 languages including English, Spanish, German, French, Japanese, and Chinese. Any voice can speak any of these languages, allowing you to produce multilingual audioguides without hiring separate voice actors for each region.
Set the language parameter to an ISO code for precise control, or let the API guess from the text script. For non-English Latin-script text, always set the language to avoid defaulting to English.
Audio files contain an inaudible watermark marking them as synthetic speech, which is standard for TTS content but ensures transparency for your audience.
- 23 languages supported
- Any voice can speak any language
- Auto-detection or manual
languageparameter
Voice Quality and Narrator Styles
Tourism audio content requires distinct vocal personalities to match the context of the experience. Our eight preset voices allow you to curate specific tones for different content types, ensuring consistent brand voice across all locations without hiring separate actors for each language.
Warm Narrator Aria for Storytelling
Aria offers a warm, clear narration style ideal for historical narratives and exhibit descriptions. Her tone remains engaging without becoming overly dramatic, making her suitable for long-form audioguides where listener fatigue is a concern.
- Optimal for historical narratives
- Clear and warm tone
- Low listener fatigue
Neutral Atlas for News-Style Updates
Atlas provides a neutral, professional delivery perfect for operational updates, weather alerts, or museum announcements. His voice conveys authority and clarity without emotional bias, ensuring information is received accurately by diverse audiences.
- Neutral delivery
- Ideal for announcements
- High clarity
Friendly Milo for Conversational Tours
Milo delivers a friendly, conversational tone that mimics a knowledgeable local guide. This style works well for walking tours or interactive museum experiences where a personal connection enhances the visitor experience.
- Conversational tone
- Local guide feel
- Engaging style
Language Support for International Tourism
The API supports 23 languages, enabling tourism startups to produce scalable audioguides for global destinations without hiring separate voice actors for each region. This coverage includes major tourist markets like Spain, France, Japan, and Germany, ensuring your audio content reaches international visitors in their native tongues.
23 Languages Available for Audioguides
Support covers 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 preset can speak any of these languages, allowing consistent audio branding across diverse locations.
This broad selection ensures that museum exhibits and hotel concierge messages can be localized efficiently, maintaining high quality without the logistical overhead of managing multiple voice talent pools.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
Auto-Detection vs. Manual Language Setting
If the language field is omitted, the API guesses the language from the text script, defaulting to English for Latin characters. For non-English Latin text, explicitly setting the language parameter ensures accurate pronunciation and avoids misinterpretation.
This flexibility allows developers to simplify integration for single-language tours while retaining precision for multilingual content.
Support for Major Tourist Markets
With support for Spanish, French, German, Japanese, and Chinese, the API meets the needs of high-traffic tourism hubs. Startups can deploy audio guides in the primary languages of their target visitor demographics without additional licensing fees.
Pricing Model for Startups
Our pricing model is built for tourism startups that need scalable audio without fixed overhead. You pay strictly for what you generate: $10 per 1 million characters, which equals $0.01 per 1,000 characters. Spaces and punctuation count toward this total, ensuring predictable costs for detailed museum descriptions or long city tour scripts.
Pay Per Character Usage
Billing is precise and proportional to your content volume. A 4,000-character audio guide costs just $0.04. If a synthesis request fails, you are not charged, so you only pay for successful audio generation.
No Monthly Fees or Minimum Spend
There are no monthly plans, seats, or minimum spend requirements. You can scale from a few audio files to thousands without contractual lock-ins. Top-ups via USDT (TRC-20) are credited automatically within minutes, allowing you to control cash flow precisely.
- No monthly fees or subscriptions
- No minimum spend requirements
- Automatic credit via USDT (TRC-20)
Transparent Cost for Long Texts
Costs remain stable regardless of text length, provided you split texts into 4,096-character chunks. The per-character rate ensures that extensive historical narratives or multi-stop walking tours do not incur disproportionate fees compared to short exhibit labels.

Technical Implementation for Developers
The API uses a standard REST endpoint compatible with the OpenAI SDK, requiring only a base URL change to https://api.a2229.info/v1. Authentication uses a Bearer token, allowing seamless integration into existing tourism tech stacks without custom wrappers.
Audio is returned in standard formats like MP3, WAV, Opus, FLAC, and PCM, ensuring broad compatibility with museum playback systems and web players.
OpenAI SDK Compatibility
Send a POST request to /v1/audio/speech with JSON containing model, input, voice, and response_format. The official OpenAI SDK works directly by setting the base_url to our API domain, simplifying development for teams already familiar with that ecosystem.
- Endpoint: POST /v1/audio/speech
- Auth: Authorization: Bearer tts-...
- SDK: Compatible with OpenAI Python/Node.js libraries
Standard Audio Formats (MP3, WAV)
Choose from MP3, WAV, Opus, FLAC, or PCM based on your infrastructure needs. MP3 offers efficient storage for web delivery, while WAV provides uncompressed quality for local museum kiosks. FLAC preserves lossless quality for archival audio guides.
Rate Limits and Queue Handling
Standard keys allow 60 requests per minute, with 10 per minute before the first top-up. Requests are processed sequentially, so high-volume tourism apps should implement client-side splitting for texts over 4096 characters. Failed requests incur no charge, protecting your budget during testing.
- 60 requests/minute per key
- Sequential processing may cause queue delays
- Split long texts manually for best performance
Data Privacy and Licensing
Input text is processed solely to generate the audio file and is not retained for model training. Generated audio is returned in the response and not stored on our servers, ensuring your tour scripts remain proprietary. Request metadata, such as timing and character counts, is kept exclusively for billing purposes.
Input Text Not Used for Training
All input text is used only to produce the audio and is not used for training.
Licensed Voice Models (LibriVox)
Our eight voice presets are based on licensed recordings from LibriTTS-R (CC BY 4.0). These models originate from LibriVox public-domain audiobook readings, providing a legally sound foundation for commercial audio guide production without additional licensing fees.
Synthetic Watermarking in Audio
Every generated audio file carries an inaudible watermark that marks it as synthetic speech. This allows for easy identification of AI-generated content in your museum exhibits or video tours, ensuring transparency with your visitors.
Common Use Cases in Tourism
Text-to-speech APIs enable tourism startups and museums to produce scalable audio content without hiring voice actors for every language. This approach ensures consistent brand voice across locations while significantly reducing production time for video tours and exhibit descriptions.
By leveraging pre-defined narrator voices, organizations can maintain professional audio quality across diverse content types, from museum guides to hotel concierge messages, without the logistical overhead of coordinating multiple recording sessions.
Museum Exhibit Descriptions
Generate concise descriptions for hundreds of artifacts using a consistent narrator voice like Aria or Luna. This eliminates the need to book studio time for each new exhibit, allowing curators to update audio guides rapidly as collections rotate.
The API handles texts up to 4,096 characters. If descriptions exceed this limit, you can split and join the audio files.
- Ideal for rotating exhibitions requiring frequent audio updates
- Consistent tone across all artifacts in a gallery
- No need for repeated studio bookings
Hotel Concierge Audio Messages
Create welcoming messages for hotel rooms or digital concierge kiosks using voices like Milo for a friendly, conversational tone. These audio files can be easily updated to reflect seasonal promotions or local event information without re-recording.
With support for 23 languages, hotels can provide localized audio instructions for international guests, ensuring clear communication regarding amenities, Wi-Fi passwords, and breakfast hours without relying on human multilingual staff for every query.
- Seasonal updates without re-recording
- Local language support for international guests
- Easy integration with digital kiosks
City Walking Tour Narration
Produce high-quality narration for self-guided walking tours that can be accessed via mobile apps or web browsers. The API’s speed settings allow for adjusted pacing, from 0.25x to 4.0x speed.
Since requests are processed sequentially, long tour routes can be split into logical segments. This allows for modular content creation, where each segment can be updated independently without affecting the entire tour audio file.
- Modular content for different tour segments
- Adjustable speed for walking pace
- Support for major tourist languages
Limitations to Consider
While the API delivers high-quality synthetic speech, it does not support streaming or WebSocket connections. Audio is generated as a complete file, meaning short phrases may take a few seconds to return, and longer texts are processed sequentially, potentially queuing under load.
The service uses fixed narrator presets rather than custom voice cloning. You cannot upload a unique voice sample for exclusive use, though the eight existing models are licensed from LibriVox and include inaudible synthetic watermarks.
No Streaming or Real-Time Audio
Audio is returned only after full synthesis completes. This batch processing ensures high fidelity but introduces latency, making it less suitable for live, interactive voice bots where sub-second response is critical.
No Custom Voice Cloning
Customers cannot upload their own voice samples for cloning. The eight available presets (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion) are the only options, though they cover diverse narrative styles.
Single Language per Request Limitation
Each request handles a single language context. While auto-detection works for Latin scripts, non-English Latin text requires explicit language codes. Mixing languages within one 4096-character input may yield inconsistent pronunciation.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. This clarity helps developers quickly prototype audio guides for museums or tourist apps without guessing parameter formats or encoding requirements.
23 languages
Any voice speaks any of these languages.
