Frequently asked questions
What makes your text-to-speech realistic for announcements?
Our voices are based on licensed recordings from LibriTTS-R (CC BY 4.0), from LibriVox public-domain audiobook readings. This foundation provides natural, human-sounding speech patterns ideal for transport environments. Each generated file also carries an inaudible watermark to clearly mark it as synthetic speech, ensuring transparency for compliance and accessibility standards.
Which voices are best for station announcements?
For clear, reliable announcements, we recommend Atlas (neutral, news-style) or Sage (calm, instructional). These presets are designed for clarity and ease of understanding in noisy environments. You can adjust the speed between 0.25 and 4.0 to ensure the audio remains crisp and legible for passengers across various transport modes.
Can I use the API for live schedule updates?
Yes, you can generate audio files on demand via our API. While typical speed is roughly real-time for long texts, short phrases may take a few seconds. Requests are processed sequentially, so under heavy load, generation might wait in a queue. We do not offer streaming or WebSocket endpoints for real-time partial audio.
Do you support multiple languages in one request?
No, each API request handles a single language. You must specify the language using an ISO code or allow our system to guess it from the script. Our API supports 23 languages, including English, Spanish, German, and Japanese, but you must manage multilingual content by sending separate requests for each language.
Is the audio file format compatible with legacy hardware?
You can choose from mp3, wav, opus, flac, or pcm formats. Standard formats like WAV and MP3 are widely supported by legacy transport announcement systems. You can select the appropriate response_format in your JSON body to ensure compatibility with your existing hardware infrastructure without needing additional conversion tools.
How do you handle non-English text with Latin script?
If the language field is not set, the system guesses the language from the script, defaulting to English for Latin text. For non-English Latin-script text, you should explicitly set the language parameter to ensure accurate pronunciation. Supported languages include Danish, Dutch, Finnish, Norwegian, Portuguese, Swedish, and others in our 23-language list.
What happens if an announcement generation 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 with the same parameters, keeping in mind your rate limit of 60 requests per minute per key.
Do you store the audio files I generate?
No, we do not store the generated audio files. The audio is returned directly in the API response and is not saved on our servers. We only retain request metadata, such as time, character count, and status, for billing purposes. Your input text is used only to produce the audio and is not used for training.
Can I upload my own voice for announcements?
No, we do not offer custom voice cloning or the ability to upload your own voice. You must choose from our 8 preset voices (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion). These voices are based on licensed recordings and are designed to provide consistent, high-quality speech for your transport announcements.
Is there a minimum spend for transport companies?
No, there is no minimum spend, no monthly plans, and no seats. You pay $0.01 per 1,000 characters. You can top up with a minimum of $5 via USDT on the TRON network. Balances do not expire, and you only pay for the characters you actually use for your announcements.
Do you offer SLA guarantees for uptime?
No, we do not offer an SLA or uptime guarantee. We also do not provide enterprise contracts, on-premise installations, or certifications like SOC 2 or HIPAA. Support is by email only.
How long does it take to generate an announcement?
Typical speed is roughly real-time for long texts, about 25 characters per second. Short phrases may take a few seconds to generate. Requests are processed sequentially, so under load, a request may wait in a queue. For very long announcements, split the text into parts of up to 4096 characters.
Can I use the API for emergency alerts?
Yes, you can generate emergency alert audio files on demand. Since each request supports up to 4096 characters, you can create concise, clear messages. You can adjust the speed and voice to ensure the alert is audible and understandable. However, note that we do not offer streaming, so you must generate the file first.
Do I need a developer to integrate this?
Our API is designed to be straightforward. It follows the same request shape as the OpenAI speech endpoint, so you can use standard HTTP requests or the OpenAI SDK by changing the base URL. We provide a ready curl example on your dashboard to get started quickly.
First request
curl
curl https://api.55gg.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 choose synthetic announcements?
Synthetic announcements ensure your transit network maintains a consistent voice across all vehicles and stations, regardless of location or time of day. Unlike human recordings, text-to-speech allows you to update messages instantly without scheduling studio sessions, ensuring passengers always hear accurate, up-to-date information.
This approach simplifies fleet-wide communication by standardizing audio output. Whether it’s a routine stop announcement or an urgent delay notification, the quality remains uniform, helping passengers recognize your service's voice even when the specific message changes.
Consistent brand voice across fleets
Standardize your transit announcements across every vehicle and station. By using a single voice preset like Atlas for news-style updates or Sage for calm instructions, you eliminate the variability of different human narrators. This consistency helps passengers quickly recognize your service, improving trust and reducing confusion during their commute.
- Uniform audio quality across all fleet vehicles
- Single voice preset for all stations and stops
- Professional tone that matches your transit brand
Instant updates for schedule changes
When delays occur or routes shift, you can update the announcement text in seconds and generate a new audio file immediately. There is no need to wait for a voice actor’s availability or re-record entire segments. This agility ensures passengers receive accurate, real-time information without delay, reducing frustration during unexpected changes.
Text inputs up to 4096 characters allow for detailed instructions, including platform numbers and transfer details. You can generate and deploy new announcements as often as needed, keeping your information current without incurring extra production costs for each update.
- Update text and generate audio in minutes
- No studio scheduling required for changes
- Support for detailed, multi-sentence announcements
Accessible compliance for all users
Clear, consistent audio announcements are essential for accessibility. Synthetic speech delivers precise enunciation and consistent volume, aiding passengers with hearing impairments or those unfamiliar with the local language. By standardizing your audio, you ensure that every passenger receives the same clear information, regardless of which vehicle they board.
Audio files in standard formats like MP3 or WAV integrate easily into existing transit PA systems. This compatibility ensures that your announcements meet accessibility standards without requiring complex infrastructure upgrades or specialized playback hardware.
- Clear enunciation aids hearing-impaired passengers
- Standard MP3/WAV formats for easy integration
- Consistent volume and tone across all announcements
Voice catalog for transit
Transport announcements require clarity and trust. Our voice presets are engineered for specific transit contexts, ensuring passengers receive consistent, human-sounding information regardless of the station or vehicle.
Atlas for neutral news-style updates
Atlas provides a neutral, news-style voice ideal for station departures and route changes. Its consistent tone ensures clarity across diverse environments, making it perfect for standard operational announcements.
Sage for calm instructional guidance
Sage offers a calm, instructional voice suitable for step-by-step guidance. Use it for detailed instructions on boarding procedures or safety protocols, ensuring passengers remain relaxed and informed.
Felix for energetic promo messages
Felix brings energy to promotional messages and special offers. Its upbeat delivery captures attention, making it ideal for advertising local events or service upgrades.
Technical reliability
Transport announcements require consistent timing across large fleets. Our API generates audio at roughly real-time speed, processing approximately 25 characters per second. Short phrases complete in seconds, while longer route descriptions scale linearly. Under load, requests queue sequentially, ensuring predictable delivery without complex infrastructure management.
We deliver standardized MP3 and WAV files compatible with legacy transit PA systems. No custom codecs or conversion pipelines are needed. Each file is ready for immediate playback on station displays and vehicle speakers. Inaudible watermarking marks every file as synthetic, aiding compliance audits without affecting audio quality.
Predictable real-time generation speeds
Generation speed averages one second of audio per 25 characters of input. Short announcements finish in seconds. Longer texts are split into 4,096-character chunks and joined by the client. Requests process sequentially, so high volume may introduce minor queue delays, but timing remains consistent and predictable.
- ~25 characters per second of generation time
- Short phrases complete in seconds
- Sequential processing ensures consistent timing
Standard MP3/WAV formats for legacy systems
We support MP3, WAV, Opus, FLAC, and PCM formats. Most transit legacy systems accept MP3 or WAV directly. This eliminates the need for intermediate conversion tools or proprietary bridges. Your existing playback infrastructure works immediately with our output files.
- MP3 and WAV support for legacy hardware
- Opus, FLAC, and PCM also available
- No conversion pipelines required
Inaudible watermarking for synthetic content
Every generated audio file carries an inaudible digital watermark. This marks the content as synthetic speech, helping you meet accessibility and safety disclosure standards. The watermark does not affect audio clarity or playback quality on any device.
- Inaudible synthetic speech watermark
- Meets disclosure compliance standards
- Zero impact on audio quality
Global reach, local clarity
Deploy announcements in 23 languages, from Arabic and Hindi to Swedish and Turkish. Every voice preset supports all listed languages, ensuring consistent quality whether you are broadcasting in English or Swahili.
23 supported languages including Arabic and Hindi
SpeakPath covers major transit hubs across diverse regions. Our 23 languages include Arabic, Hindi, Chinese, Japanese, and Spanish, allowing you to serve global passenger bases without managing multiple voice providers.
- 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 for Latin-script defaults
If you do not specify a language, the API guesses the language based on the text script. For Latin-script text, it defaults to English, which works well for most English-speaking transit networks.
Explicit language codes for non-English text
For non-English Latin text, always set the language parameter to ensure accurate pronunciation. This prevents the engine from defaulting to English when processing words like 'café' or 'naïve'.
Simple integration
Integrate SpeakPath into your transport announcement pipeline without learning a new SDK. Our API follows the standard OpenAI speech endpoint structure, allowing you to use familiar OpenAI client libraries by simply pointing them to our base URL. This reduces integration time and leverages existing developer knowledge within your product team.
Send a standard JSON payload with your text, selected voice, and desired format. The system returns a standardized audio file ready for playback on station displays or vehicle infotainment systems, ensuring consistent delivery across your entire fleet.
OpenAI SDK compatible endpoint
Use the official OpenAI SDKs by setting the base URL to https://api.55gg.info/v1. This compatibility means your existing authentication and request logic works immediately, minimizing friction during deployment.
The endpoint accepts standard POST requests to /v1/audio/speech, handling the heavy lifting of synthesis while you focus on integrating the audio into your existing broadcast workflows.
- Base URL: https://api.55gg.info/v1
- Endpoint: POST /v1/audio/speech
- Supports official OpenAI SDKs
Standard JSON request body
Construct your request with simple fields: model, input text, voice, and response format. You can also specify speed and language for precise control over the output characteristics.
Ensure your input text stays within the 4096 character limit per request. For longer announcements, split the text and join the resulting audio files seamlessly.
- Fields: model, input, voice, response_format, speed
- Max 4096 characters per request
- Supports MP3, WAV, OPUS, FLAC, PCM
No custom SDKs required
You do not need to install or maintain a proprietary library. Use plain HTTP requests or any standard JSON-compatible client to interact with the API.
This approach ensures long-term stability and flexibility, as your integration relies on universal standards rather than version-specific SDKs that may change or break.
- Use any HTTP client
- No proprietary dependencies
- Standard JSON format

Pricing for scale
SpeakPath charges a flat rate of $10 per million characters, which breaks down to $0.01 per 1,000 characters. This transparent model means you only pay for what you actually use, with no hidden fees or surprise overages for high-volume announcement deployments.
Your balance never expires, and failed requests are free. You can top up instantly using USDT (TRC-20) on the TRON network, with a minimum deposit of just $5. This approach eliminates the need for complex enterprise contracts or waiting for monthly invoices.
Pay only for successful generations
If synthesis fails, the charge is returned in full. You are not billed for errors, ensuring your costs directly reflect usable audio files delivered to your fleet systems.
No monthly minimums or seats
There are no recurring subscriptions, seat licenses, or minimum spend requirements. Whether you are generating ten announcements a day or ten thousand, the pricing structure remains consistent and predictable.
Transparent character-based billing
Billing is based strictly on character count, including spaces and punctuation. With a free trial credit of $0.50 (valid for 7 days), you can test the system immediately upon signing in with Google. Accounts created with email and password start without trial credit.
- Rate: $10 per 1 million characters
- Min. Top-up: $5 via USDT (TRC-20)
- Free trial: $0.50 with Google sign-in
Data privacy
Transport announcements often contain destination names, schedule changes, or safety instructions that might be considered sensitive operational data. SpeakPath ensures that your input text is used exclusively for the immediate synthesis of audio and is not retained for model training, keeping your proprietary route data secure.
Once the audio file is generated and returned in the API response, it is not stored on our servers. This minimizes data exposure and ensures that your announcement content remains ephemeral unless you choose to save it in your own infrastructure.
Input text not used for training
Unlike some consumer-grade services, we do not use your announcement text to improve our voice models. Your input is processed solely to generate the requested audio file, ensuring that specific transit information remains private to your organization.
Audio returned in response, not stored
The generated audio file is delivered directly in the response payload. We do not archive these files on our servers after delivery. This approach reduces storage costs and limits the window where your audio data exists outside your control.
Metadata kept for billing only
We retain minimal request metadata, including timestamp, character count, and status, strictly for accurate billing and troubleshooting. This data does not include the content of your announcements and is used only to verify usage against your account balance.
Get started today
Start generating transport announcements immediately with a $0.50 free trial available when you sign in via Google. This credit covers approximately 50,000 characters and remains valid for seven days, allowing you to validate voice quality and integration speed without upfront commitment.
Once your account is active, you receive an API key instantly. This key grants access to the full API endpoint, enabling you to begin testing voice presets like Atlas or Sage right away with zero configuration overhead.
Free trial with Google sign-in
Sign up using your Google account to receive an automatic $0.50 credit. This trial lasts for 7 days and allows you to generate roughly 50,000 characters of audio. It is the fastest way to verify that the synthetic voices meet your fleet’s accessibility standards before committing to top-ups.
Email/password accounts do not include this trial credit. No credit card or minimum spend is required to start.
- Google sign-in only
- $0.50 automatic credit
- 7-day validity
- Approx. 50,000 characters
Instant API key generation
Your API key is generated immediately upon account creation and is visible on the dashboard. You can use this key to authenticate requests to the endpoint without waiting for manual approval.
If needed, you can issue a new key at any time. The previous key stops working instantly, ensuring secure access management across your product team.
Use the key in the Authorization header as a Bearer token. The endpoint accepts standard JSON bodies compatible with the OpenAI SDK.
- Immediate availability
- No manual approval
- Revocable instantly
Documented curl examples
The dashboard provides ready-to-use curl commands for common scenarios. You can copy these directly into your terminal to test voice generation with presets like Atlas or Luna.
These examples demonstrate the standard JSON structure, including model selection, voice parameters, and input text limits. They serve as a reliable reference for your engineering team during integration.
Ensure your base URL is set to https://api.55gg.info/v1 to match the documented endpoints.
- Copy-paste ready
- JSON structure shown
- Endpoint verified
Support and limits
SpeakPath provides straightforward email support for technical queries. We do not offer live chat or phone lines, so please include your API key and request details in your message for faster resolution.
Our infrastructure is designed for reliability, but specific operational limits ensure fair usage across the network. Understanding these constraints helps you plan your announcement workflows effectively.
When integrating SpeakPath, remember that requests are processed sequentially. High volume may result in brief queuing during peak times, so designing your application to handle retries gracefully is recommended.
Email support for technical queries
We do not offer live chat or phone lines. Please include your API key and request details in your message for faster resolution.
Support is handled entirely via email, allowing for detailed technical context without the pressure of real-time conversation.
60 requests per minute rate limit
Each API key is limited to 60 requests per minute. This ensures stable processing for your announcement generation.
Before your first top-up, the limit is 10 requests per minute. This lower threshold allows you to test integration stability without consuming significant credit.
4096 character input limit per request
Each request can process up to 4,096 characters of text. Longer announcements should be split into parts and joined for seamless playback.
Spaces and punctuation count toward the character limit. Plan your text structure accordingly to maximize efficiency within each request.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Pay attention to passenger comments regarding audio clarity and volume.
23 languages
Any voice speaks any of these languages.
