Frequently asked questions
What is this TTS API?
VoxMuse is a cloud text-to-speech API where you send text and receive an audio file. It uses the standard OpenAI request shape, so you can use official OpenAI SDKs by changing the base URL to https://api.getchickensubway.com/v1. You authenticate with a Bearer token and receive synthesized audio in formats like MP3 or WAV.
Is this affiliated with OpenAI?
No, VoxMuse is not affiliated with OpenAI. We use a compatible endpoint structure, so the official OpenAI SDKs work by changing the base URL. Our voices are based on licensed recordings from LibriTTS-R and LibriVox. We are an independent storefront providing a character-based TTS API for indie developers without OpenAI branding or infrastructure.
Can I use this for commercial tourism apps?
Yes, this API is designed for indie developers building audioguides and tourism experiences. There are no enterprise contracts, minimum spends, or seat limits. You simply pay for the characters you generate. The API supports standard audio formats and integrates easily into your existing workflow using standard HTTP requests or SDKs.
How many languages are supported?
We support 23 languages: 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. If you do not set the language, it is guessed from the script.
Do you support SSML for pacing?
No, we do not support SSML. The API accepts plain text input up to 4096 characters. You can control pacing using the speed parameter, which accepts values between 0.25 and 4.0. For longer texts, split your input into parts and join the resulting audio files manually in your application.
What happens if a request fails?
If synthesis fails, the charge is returned in full. You are only billed for successful character generation. Failed requests do not consume your balance. You can monitor status codes in the response to determine if a request succeeded or failed, allowing you to retry or handle errors appropriately in your app.
Can I clone my own voice?
No, we do not offer custom voice cloning or uploading your own voice. We provide eight preset voices based on licensed recordings from LibriTTS-R and LibriVox. These include female voices like Aria and Nova, and male voices like Felix and Milo. Each generated file carries an inaudible watermark marking it as synthetic speech.
Is there a monthly subscription?
No, there are no monthly plans, subscriptions, or seats. You pay only for what you generate at $10 per 1 million characters. There is no minimum spend. Your balance does not expire. You can start with a free trial of $0.50 credit when signing in with Google, valid for 7 days.
How do I pay for the API?
We accept USDT on the TRON network (TRC-20) only. The minimum top-up is $5, and credits are added automatically within minutes after transfer. We do not accept cards, PayPal, or offer invoices with VAT. You can manage your account and view your balance on the /get-api-key/ page after signing up.
What audio formats are available?
You can request audio in mp3, wav, opus, flac, or pcm formats. Specify the response_format field in your JSON body. These formats cover most use cases for tourism apps and audioguides. The API returns the audio file directly in the response, allowing you to save it as needed.
Are the voices realistic?
Our voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and LibriVox public-domain audiobook readings. They include warm narrators, upbeat presenters, and calm instructional tones. Every generated file carries an inaudible watermark that marks it as synthetic speech. You can choose from eight presets to fit your tourism content style.
Do you store my generated audio?
No, generated audio is returned in the response and not stored. Input text is used only to produce the audio and is not used for training. We keep request metadata, such as time, characters, and status, for billing purposes. Your data privacy is maintained without retaining the actual audio files on our servers.
How fast is the generation?
Typical speed is roughly real time, generating about 25 characters of text per second of audio. Short phrases may take a few seconds. Requests are processed one after another, so under load, a request may wait in a queue. Longer texts should be split into parts of up to 4096 characters.
Is there an SLA for uptime?
No, we do not provide an SLA or uptime guarantee. We are an independent storefront without enterprise contracts or certifications like SOC 2 or HIPAA. Support is available by email only. There is no phone number or live chat. We aim for reliability but do not promise specific availability percentages.
First request
curl
curl https://api.getchickensubway.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 choose this API for tourism apps?
Tourism apps require consistent audio budgets across varying content lengths. Our API charges strictly by character count ($10 per million characters), eliminating unexpected overages from variable speech durations. Failed requests are free, ensuring you only pay for usable audio.
Support 23 languages including English, Spanish, French, German, Japanese, and Arabic. This coverage allows indie developers to serve global travelers without managing multiple voice providers or negotiating separate contracts for each region.
Predictable character-based pricing
Chargebacks are rare in tourism, but unpredictable costs can break small budgets. By pricing at $0.01 per 1,000 characters, you can estimate audio costs before generation. Spaces and punctuation count as characters, making calculations straightforward.
- $10 per 1 million characters
- $0.01 per 1,000 characters
- Failed requests are free
23 languages for global travelers
Cover major markets without complex integrations. Our 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. Any voice can speak any of these languages.
No monthly minimums or seats
Start with a $0.50 trial credit valid for 7 days. Top up via USDT (TRC-20) with a $5 minimum. No subscriptions, no invoices, and no seats to manage. Pay only for what you generate.
Voice presets for different guide styles
We match voice personalities to your content type, ensuring the audio fits the context. Our eight presets cover the spectrum from deep documentary tones to bright, upbeat narration, allowing you to maintain consistent branding across your tourism app without hiring voice actors.
Warm narrators for history tours
Aria provides a warm, clear narrative voice ideal for historical storytelling. Luna offers a soft, storytelling tone that suits atmospheric museum guides. These voices help create an engaging, immersive experience for visitors exploring heritage sites.
- Aria: Warm, clear narrator
- Luna: Soft, storytelling tone
Energetic voices for promo spots
Felix delivers an energetic, promo-ready voice perfect for short advertisements or highlights. Milo provides a friendly, conversational tone suitable for casual introductions. These voices keep the energy high in promotional sections of your app.
- Felix: Energetic, promo-focused
- Milo: Friendly, conversational
Calm tones for instructional guides
Sage offers a calm, instructional voice ideal for wayfinding or technical instructions. Atlas provides a neutral, news-like tone for factual data delivery. Orion’s deep, documentary style works well for serious, detailed content. These voices ensure clarity and reduce listener fatigue during longer tours.
- Sage: Calm, instructional
- Atlas: Neutral, news-style
- Orion: Deep, documentary
Technical integration for developers
Integrate quickly using standard tools. The API endpoint matches the OpenAI speech format, allowing you to use the official OpenAI SDKs by simply changing the base_url to https://api.getchickensubway.com/v1. This means your existing code works with minimal adjustment, avoiding the need for custom client libraries.
You can issue requests using any standard HTTP client. The service accepts JSON bodies and returns audio files directly, making it easy to embed synthesized speech into mobile apps or web-based audioguides without complex middleware.
OpenAI SDK compatible endpoint
The API uses the standard OpenAI request shape. By pointing your SDK to https://api.getchickensubway.com/v1, you can use familiar methods to generate speech. This compatibility reduces boilerplate code and lets you leverage existing client libraries for authentication and retries.
Authentication is handled via a Bearer token in the Authorization header. You receive a single active key per account, which can be rotated instantly if needed, ensuring secure access for your tourism applications.
- Endpoint:
POST https://api.getchickensubway.com/v1/audio/speech - SDK compatibility: Official OpenAI SDKs work with
base_urladjustment - Auth:
Authorization: Bearer tts-...
Multiple audio formats (mp3, wav, opus)
Choose the output format that best fits your deployment needs. Supported formats include mp3, wav, opus, flac, and pcm. This flexibility allows you to balance file size and quality depending on whether you are storing audio for offline museum guides or using it in playback.
Using opus or mp3 reduces bandwidth usage, which is crucial for users with limited data plans. WAV or FLAC provides lossless quality for archival purposes or high-end audio installations.
- Formats: mp3, wav, opus, flac, pcm
- Optimize for bandwidth or quality
- No streaming or WebSocket support
Simple JSON request structure
Requests are straightforward JSON objects. You provide the model (tts-1 or tts-1-hd), input text, voice preset, and desired response_format. The language field is optional; if omitted, the system guesses the language from the script, defaulting to English for Latin text.
Example request fields include speed (0.25–4.0) and voice (e.g., 'aria' for warm narration). The API returns the audio file directly in the response, allowing immediate playback or storage in your application’s asset library.
- Fields:
model,input,voice,response_format,speed,language - Max input: 4096 characters
- No SSML or phoneme dictionaries
Quality and licensing
We use voice recordings from the LibriVox public-domain collection, licensed under CC BY 4.0. These authentic, human-read audiobook samples provide a natural foundation for our synthesis engine, ensuring high-quality output without the need for complex neural training on proprietary datasets.
Licensed recordings from LibriVox
Our voices are derived from the LibriTTS-R dataset, sourced from LibriVox public-domain audiobook readings. This approach leverages real human performance rather than purely synthetic generation, giving your museum guides a grounded, realistic tone.
Synthetic watermark on all files
Every audio file generated by our API includes an inaudible watermark. This metadata marks the file as synthetic speech, which helps maintain transparency when distributing audioguides or tourism content to end-users.
No training on your input text
Your input text is used solely to produce the immediate audio response and is not stored for model training. We retain only request metadata (time, characters, status) for billing purposes, ensuring your content remains private.
Supporting long-form content
Tourism apps often require longer narratives, such as full museum tours or city walking guides. This API handles large texts by splitting them into segments of up to 4096 characters, ensuring reliable processing without timeouts.
The service joins the resulting audio files seamlessly, maintaining a consistent voice and tone throughout the entire experience. This allows you to deliver complete, coherent audio guides without manual editing of the audio files.
Split and join large texts automatically
When your input exceeds 4096 characters, the system divides the text into manageable chunks. Each chunk is synthesized individually, and the audio files are concatenated in order. This process happens automatically, so you receive a single, continuous audio file ready for playback.
Consistent voice across segments
Since all segments are generated using the same voice preset and speed settings, the audio remains uniform. There are no jarring transitions between parts, which is essential for maintaining immersion in audioguides and narrative tours.
Handle up to 4096 characters per request
Each API request accepts a maximum of 4096 characters, including spaces and punctuation. This limit ensures stable processing times. For texts longer than this, simply split your content and send multiple requests, or use the automatic splitting feature if available in your client.
- Character limit includes spaces and punctuation
- Ideal for short exhibits or multi-stop tours
- Prevents processing timeouts for long documents

Deployment flexibility
VoxMuse is a cloud-based API, meaning you integrate directly without provisioning servers or managing infrastructure. Your tourism app or audioguide simply sends text to our endpoint and receives an audio file, keeping your development stack lightweight and focused on content rather than backend maintenance.
Cloud-based API, no server setup
By offloading synthesis to our cloud, you avoid the complexity of hosting TTS engines. This allows indie developers to deploy updates quickly without worrying about server scaling.
Works with standard HTTP clients
You can use any standard HTTP client or the OpenAI SDK (with a adjusted base URL) to send requests. The simple JSON structure mirrors familiar patterns, so you can start generating audio with minimal code changes or new library dependencies.
Scalable for small to medium tours
The API handles variable traffic, making it ideal for indie projects that may see spikes during peak tourist seasons. With no minimum spend or complex tiered plans, you only pay for the characters your tours actually generate.
Cost efficiency for indie projects
TTS API for indie tourism projects avoids complex subscriptions and enterprise contracts. You pay only for what you generate, making it easy to budget for small museum guides or local audio tours without worrying about minimum monthly fees or per-seat charges.
Failed synthesis requests are free, so you only pay for successful audio generation. This protects your budget during testing or when handling variable input lengths.
$10 per million characters
Pricing is straightforward: $10 per 1 million characters, or $0.01 per 1,000 characters. Spaces and punctuation count toward the limit. There are no monthly plans, no seats, and no minimum spend.
This model works well for indie developers building audioguides, as costs scale directly with content volume.
- No monthly subscriptions
- No per-seat fees
- Transparent character-based billing
Free failed requests
If synthesis fails, the charge is returned in full. You only pay for successfully generated audio files.
This ensures you are not penalized for transient errors or invalid inputs.
Start with $0.50 trial credit
New accounts signed up with Google receive $0.50 of trial credit (about 50,000 characters) valid for 7 days. This allows you to test integration before committing funds.
Email/password accounts start without trial credit.
Language coverage for museums
Tourists expect content in their native tongue. Our API supports 23 languages, covering major European and Asian markets including English, Spanish, French, German, Japanese, and Chinese. This allows indie developers to reach global audiences without managing multiple voice providers.
We also support Arabic and Hebrew for Middle Eastern tourism routes. Non-English Latin text requires explicit language setting to ensure correct pronunciation, as default detection assumes English. Vietnamese, Thai, and Ukrainian are not yet supported.
Major European and Asian languages
Covering 23 languages ensures your audioguide reaches a broad audience. Supported languages include 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.
- 23 supported languages for global tourism
Arabic and Hebrew support
We include Arabic and Hebrew to serve Middle Eastern heritage sites and tours. Right-to-left script support is handled via standard ISO codes in the request body.
Auto-detection for Latin scripts
If the language field is omitted, the system guesses the language from the script. Latin-script text defaults to English. For other Latin-based languages like French or German, you must explicitly set the language parameter to ensure accurate pronunciation.
Developer experience
The API mirrors the OpenAI speech endpoint structure, meaning you can reuse existing code by simply updating the base_url to https://api.getchickensubway.com/v1. This compatibility allows indie developers to integrate standard tools and SDKs without learning a proprietary protocol.
We provide ready-to-use curl examples and clear documentation to help you test the integration quickly. The single JSON request structure handles model selection, voice choice, and audio format in one call, keeping your implementation clean and straightforward.
Clear documentation and examples
Our documentation includes practical curl examples tailored for tourism use cases. You can quickly verify the endpoint behavior before committing to full integration.
The request structure is minimal: specify the model, input text, voice, and desired format. This simplicity reduces debugging time for developers building audioguide apps.
- OpenAI SDK compatible endpoint
- Ready curl examples
- Minimal JSON structure
Single API key management
Each account uses one active API key, simplifying authentication in your codebase. You do not need to manage multiple tokens for different environments or users.
The key is displayed on your dashboard alongside your balance and usage statistics, providing immediate visibility into your project's consumption.
Immediate key rotation support
If you suspect a key leak, you can issue a new key instantly. The old key stops working immediately, ensuring your audio generation remains secure without complex rotation logic.
This feature is available directly from the /get-api-key/ page, allowing quick response to security concerns during development or production.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. We build immersive audio stories for independent creators who want to bring museums to life through sound.
23 languages
Any voice speaks any of these languages.
Sources and further reading
- Wikipedia
- IETF RFC 7845: Ogg Opus
- curl documentation
- Node.js documentation
- Python documentation
- LibriTTS-R corpus (OpenSLR 141)
- LibriTTS-R paper (arXiv 2305.18802)
- W3C WAI: making audio and video accessible
- LibriVox public-domain audiobooks
- Creative Commons Attribution 4.0
- ISO 639 language codes
- FFmpeg documentation
