Frequently asked questions
How do I set the base_url for the OpenAI SDK?
Set the base_url to https://api.99dtpix.com/v1. This allows you to use the official OpenAI SDKs by simply changing the endpoint configuration. Your requests will route correctly to the VoxRoute API, enabling integration with your existing transport software stacks.
What is the character limit per request?
Each request supports up to 4096 characters of input text. If your announcement or dashboard message exceeds this limit, you must split the text into smaller parts. Generate audio for each segment separately and then join the resulting audio files to create the complete output.
Can I use the API without the OpenAI SDK?
Yes. The API accepts standard JSON POST requests to the /v1/audio/speech endpoint. You can use any HTTP client library, including curl or native code, by sending the required fields like model, input, voice, and response_format. The OpenAI SDK is just a convenient wrapper, not a requirement.
How do I handle rate limits in my code?
The limit is 60 requests per minute per key. Implement exponential backoff or a simple queue in your application logic to handle 429 errors. Since requests are processed sequentially, under load requests may wait in a queue. Monitor your usage via the dashboard to ensure your transport system stays within limits.
What formats are supported for audio output?
You can request audio in mp3, wav, opus, flac, or pcm formats. Select the format that best fits your in-vehicle announcement system or logistics dashboard. The API returns the audio file directly in the response, allowing immediate playback or storage without additional conversion steps.
Does the API support streaming audio?
No. The API returns the complete audio file in the response body. There is no WebSocket or streaming endpoint for partial audio chunks. You must wait for the full synthesis to complete before playing the audio, which typically takes a few seconds for short phrases.
How do I split text longer than 4096 characters?
Split your input text into chunks of up to 4096 characters. Send each chunk as a separate POST request to the speech endpoint. Once you receive the audio files for each part, concatenate them in your application to play the full announcement or message sequence.
What happens if I exceed the rate limit?
Requests will be rejected with a 429 status code. You must wait until the rate window resets or implement a queue in your code. Failed requests are free, so no credit is deducted if the synthesis fails. Retry with backoff.
Is the API key secure enough for production?
Yes. Use the Bearer token in the Authorization header for each request. You can issue a new key at any time, which immediately invalidates the old one. This allows you to rotate keys securely without downtime, ensuring your transport systems maintain secure access to the TTS service.
Can I use the same key for multiple apps?
Yes. One active key per account can be used across multiple applications or devices. The rate limit of 60 requests per minute applies to the key as a whole. Monitor your total usage in the dashboard to ensure all your apps stay within the allowed request volume.
Do I need to specify the language for English text?
No. If you send Latin-script text without specifying the language, the API defaults to English. However, for non-English Latin-script text, you must set the language parameter explicitly. This ensures accurate pronunciation for languages like German, Dutch, or Portuguese, which might otherwise be misinterpreted as English.
How do I test the API before going live?
Sign up with Google to receive $0.50 of free trial credit. This allows about 50,000 characters of testing. Use the curl example on the dashboard to verify integration. Check the voices and formats available via the /v1/voices endpoint before deploying to your in-vehicle or dashboard systems.
Are there SDKs for mobile platforms?
No. There is no official mobile SDK, browser extension, or WordPress plugin. You can use the standard HTTP API from any mobile platform by sending JSON POST requests. The OpenAI SDK works if you adjust the base_url.
Does the API support SSML tags?
No. The API does not support SSML, phoneme dictionaries, or custom pronunciation lexicons. You control pacing and tone by adjusting the speed parameter between 0.25 and 4.0. For more precise control, split your text and adjust the voice selection or speed for each segment.
First request
curl
curl https://api.99dtpix.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.mp3Python (openai)
from openai import OpenAI
client = OpenAI(base_url="https://api.99dtpix.com/v1", api_key="tts-...")
with client.audio.speech.with_streaming_response.create(
model="tts-1", voice="nova", input="Hello from Python.", response_format="mp3"
) as response:
response.stream_to_file("hello.mp3")Node.js (openai)
import OpenAI from "openai";
import fs from "node:fs";
const client = new OpenAI({ baseURL: "https://api.99dtpix.com/v1", apiKey: process.env.TTS_API_KEY });
const res = await client.audio.speech.create({ model: "tts-1", voice: "milo", input: "Hello from Node.js." });
fs.writeFileSync("hello.mp3", Buffer.from(await res.arrayBuffer()));| voice | |
|---|---|
aria | female |
nova | female |
luna | female |
sage | female |
felix | male |
milo | male |
atlas | male |
orion | male |
| language | |
|---|---|
ar | العربية |
zh | 中文 |
da | Dansk |
nl | Nederlands |
en | English |
fi | Suomi |
fr | Français |
de | Deutsch |
el | Ελληνικά |
he | עברית |
hi | हिन्दी |
it | Italiano |
ja | 日本語 |
ko | 한국어 |
ms | Bahasa Melayu |
no | Norsk |
pl | Polski |
pt | Português |
ru | Русский |
es | Español |
sw | Kiswahili |
sv | Svenska |
tr | Türkçe |
Get Your API Key
Access the API key management interface at /get-api-key/ to begin integration. The dashboard displays your active key, current balance, and usage metrics specific to your VoxRoute account.
Sign Up on /get-api-key/
Navigate to the registration page to create your independent storefront account. No enterprise contracts or minimum spend requirements apply.
Google or Email Authentication
You may authenticate via Google OAuth or standard email and password. Google sign-ups receive a $0.50 free trial credit valid for 7 days. Email-only accounts start with zero balance and require an immediate top-up.
- Google: Instant credit, 7-day validity
- Email: No trial credit included
Key Rotation Policy
Each account holds one active API key. You can issue a new key at any time via the dashboard. The previous key becomes invalid immediately upon generation, ensuring secure transport announcements without downtime.
First Request: cURL Example
Send a POST request to https://api.99dtpix.com/v1/audio/speech with a valid Bearer token. The API accepts JSON payloads and returns binary audio data directly, allowing seamless integration into transport announcement systems without complex streaming setups.
This endpoint mirrors the OpenAI speech structure, ensuring compatibility with existing SDKs. You define the model, voice, and input text to generate the final audio file in one synchronous call.
POST Endpoint Structure
Direct your HTTP client to POST https://api.99dtpix.com/v1/audio/speech. Include the Authorization: Bearer tts-... header with your unique key from /get-api-key/. The request body must be formatted as JSON.
JSON Body Parameters
The payload requires model ("tts-1" or "tts-1-hd"), input (text up to 4096 characters), voice (one of 8 presets), and response_format (mp3, wav, opus, flac, pcm). Optional fields include speed (0.25–4.0) and language (ISO code).
model: "tts-1" or "tts-1-hd"input: string, max 4096 charsvoice: e.g., "aria", "atlas"response_format: mp3, wav, opus, flac, pcm
Response Audio File
On success, the API returns the audio binary directly in the response body. No intermediate file storage is required; your logistics dashboard can immediately play or cache the returned audio file for in-vehicle announcements.

OpenAI SDK Integration
The VoxRoute API mirrors the OpenAI speech endpoint structure, allowing you to reuse existing codebases with minimal friction. By simply redirecting the SDK client to our base URL, you can generate transport announcements or dashboard audio using familiar methods without learning a new library.
Change base_url to api.99dtpix.com
Configure your OpenAI SDK client to point to https://api.99dtpix.com/v1. This single change routes all standard speech synthesis requests to our infrastructure while maintaining compatibility with the official SDK interface.
Use Official SDK Methods
Call the standard audio.speech.create() method as you would for OpenAI. Pass your text, desired voice preset, and output format directly into the method. The SDK handles serialization and response parsing exactly as expected for standard TTS endpoints.
Authentication Header
Set the Authorization header to Bearer tts-... using your unique API key. This token authenticates your requests and tracks usage against your account balance. Ensure your key is active and has sufficient credits before initiating synthesis jobs.
Request Parameters
The API accepts two model identifiers, "tts-1" and "tts-1-hd," which utilize the same underlying engine and pricing tier. Your choice depends solely on whether you prefer standard or high-definition audio output, with no difference in cost. Selecting the correct voice preset is critical for matching the tone of your transport announcements. The API supports eight distinct presets, ranging from warm narrators to neutral news readers, all capable of speaking any of the 23 supported languages without requiring separate model calls.
Model Selection (tts-1/tts-1-hd)
Both model identifiers route through the identical synthesis engine, ensuring consistent voice quality and character limits. The primary distinction lies in the audio fidelity; "tts-1-hd" provides higher definition output suitable for clear in-vehicle announcements, while "tts-1" offers standard quality. Both are billed at the same rate of $10 per million characters.
- Use "tts-1" for standard quality announcements.
- Use "tts-1-hd" for high-definition clarity.
- Both models share identical pricing and limits.
Voice Preset List
Choose from eight licensed voice presets optimized for different announcement styles. Female voices include Aria (warm narrator), Nova (bright), Luna (soft storyteller), and Sage (calm instructor). Male options feature Felix (energetic promo), Milo (friendly), Atlas (neutral news), and Orion (deep documentary).
Each voice supports all 23 languages. You can specify the language via the optional language parameter to ensure accurate pronunciation, though the API will attempt to guess the language based on the input script if omitted.
- Aria: Warm, clear narrator
- Nova: Bright, upbeat
- Luna: Soft, storytelling
- Sage: Calm, instructional
- Felix: Energetic, promo
- Milo: Friendly, conversational
- Atlas: Neutral, news
- Orion: Deep, documentary
Response Format Options
Specify the output audio format using the response_format field. Available options include mp3, wav, opus, flac, and pcm. Choose mp3 or opus for efficient bandwidth usage in logistics dashboards, or wav and flac for higher fidelity storage. The format choice does not affect the character limit or pricing.
- mp3: Standard compressed audio
- wav: Uncompressed, high quality
- opus: Efficient compression
- flac: Lossless compression
- pcm: Raw uncompressed data
Handling Long Texts
The API accepts a maximum of 4096 characters per single request. Any input exceeding this threshold will be rejected, requiring you to segment your text into manageable chunks before sending them to the endpoint.
To handle longer announcements, such as detailed bus route descriptions or logistics updates, you must split the input text into segments of up to 4096 characters. Each segment is then submitted as a separate API call to ensure valid processing.
4096 Character Limit
Each POST request to /v1/audio/speech supports a maximum of 4096 characters in the input field. This limit includes all spaces and punctuation marks. If your text exceeds this size, the API returns an error, so segmentation is required before submission.
This constraint applies uniformly across both tts-1 and tts-1-hd models. Ensure your client-side logic calculates character counts accurately, as every space and comma counts toward the 4096-character ceiling.
- Maximum input length: 4096 characters per request.
- Includes all spaces, punctuation, and special characters.
- Exceeding the limit results in an immediate rejection.
Split and Join Strategy
Divide your long text into segments that fit within the 4096-character limit. Send each segment as a distinct POST request to the speech endpoint. Maintain the order of segments to preserve the sequence of the final audio output.
Since requests are processed sequentially under load, longer texts may experience queue delays. Ensure your application handles these asynchronous responses and buffers the audio data before concatenation to maintain playback integrity.
Audio Concatenation
The API returns individual audio files for each request. You must concatenate these files in the same order as the original text segments. This can be done using standard audio libraries in your preferred language or a simple client-side script.
Use the response_format you selected (mp3, wav, opus, etc.) to ensure compatibility with your transport dashboard or in-vehicle announcement system. The concatenated result should play seamlessly without gaps, provided you handle the audio files correctly.

Rate Limits
VoxRoute enforces a strict rate limit of 60 requests per minute per API key. This ensures stable performance for in-vehicle announcements and logistics dashboards, preventing sudden bursts from overwhelming the synthesis engine.
60 Requests per Minute
After the first top-up, you can process up to 60 synthesis requests per minute. This capacity is sufficient for most batch-processing workflows, such as pre-generating bus stop announcements or route updates.
10 Requests Before Top-Up
New accounts or those without recent top-ups start with a reduced limit of 10 requests per minute. This ensures basic usability while allowing time for USDT (TRC-20) top-ups to process, which typically take minutes.
Queue Behavior
Requests are processed sequentially. Under load, subsequent requests may wait in a queue before synthesis begins. For long texts, this delay is negligible as generation speed is roughly real-time, but rapid-fire short requests may experience slight latency during high traffic.
Error Handling
If synthesis fails, the charge is returned in full, ensuring you only pay for successful audio generation. This applies to transient errors where the API cannot produce the file despite valid parameters.
Understanding response codes helps distinguish between temporary network issues and permanent configuration errors, allowing for precise debugging of your transport announcement integration.
Failed Request Refund
Failed requests are free. If the synthesis process fails for any reason, the credit is returned to your balance immediately. This ensures accurate billing for your logistics dashboard.
Common Error Codes
Standard HTTP status codes indicate the request state. A 400 error usually means invalid JSON or missing required fields like 'model' or 'input'. A 401 error indicates an invalid or expired API key.
Rate limit errors (429) occur when you exceed 60 requests per minute. Ensure your vehicle announcement system handles these gracefully without overwhelming the queue.
Retry Logic
Implement exponential backoff for transient errors (5xx server errors). Since requests are processed sequentially, under load a request may wait in a queue. Respect the 429 response by waiting before retrying.

Voices and Languages
Select from eight licensed voices or 23 languages to match your transport system’s tone. All voices support every listed language, ensuring consistent quality across routes and dashboards.
Use warm narrators for station announcements or calm instructional tones for logistics updates. Each audio file includes an inaudible synthetic watermark for verification.
8 Preset Voices
Choose from four female and four male presets based on licensed LibriTTS-R recordings.
Options range from upbeat promo styles to neutral news delivery and deep documentary tones.
- Aria: warm, clear narrator
- Nova: bright, upbeat
- Luna: soft, storytelling
- Sage: calm, instructional
- Felix: energetic, promo
- Milo: friendly, conversational
- Atlas: neutral, news
- Orion: deep, documentary
23 Supported Languages
The API supports 23 languages, including Arabic, Chinese, English, French, German, Hindi, Japanese, Korean, Spanish, and Russian.
All voices can speak any of these languages, providing flexibility for multi-regional transport systems.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
Language Guessing Logic
If the language field is omitted, the API guesses the language from the script.
Latin script defaults to English; specify language for other Latin-script texts to avoid misidentification.
Not supported: Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, Bengali.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Welcome to the developer hub, where your creative vision meets the stage.
23 languages
Any voice speaks any of these languages.
Sources and further reading
- Wikipedia
- Python documentation
- openai-node on GitHub
- ISO 639 language codes
- LibriTTS-R paper (arXiv 2305.18802)
- openai-python on GitHub
- Creative Commons Attribution 4.0
- LibriVox public-domain audiobooks
- MDN: Web Speech API
- IETF RFC 7845: Ogg Opus
- W3C WAI: making audio and video accessible
- LibriTTS-R corpus (OpenSLR 141)
