Frequently asked questions
How do I get an API key?
Sign up on /get-api-key/ using Google or email. Your key starts with 'tts-'. One active key per account; issuing a new one stops the old one immediately. Failed requests are free, and you receive $0.50 trial credit for Google sign-ups.
What is the base URL for the API?
The base URL is https://api.zbus.info/v1. Use POST https://api.zbus.info/v1/audio/speech for synthesis. Other endpoints include GET /v1/voices and GET /v1/models. Send JSON bodies with your Bearer token in the Authorization header.
Can I use the OpenAI SDK?
Yes. Change the SDK’s base_url to https://api.zbus.info/v1 and use your site’s key. The request shape matches the OpenAI speech endpoint exactly. This works with official OpenAI SDKs without needing a custom library from us.
What voices are available?
Eight presets exist: aria, nova, luna, sage, felix, milo, atlas, and orion. Voices are based on licensed LibriTTS-R recordings. Each voice can speak any of the 23 supported languages. Generated files contain an inaudible synthetic watermark.
How many characters can I send per request?
You can send up to 4096 characters per request. For longer texts, split the input into parts of this size and join the resulting audio files. Spaces and punctuation count toward the character limit.
What audio formats are supported?
Supported formats are mp3, wav, opus, flac, and pcm. Set the response_format field in your JSON body. All formats are returned directly in the API response without additional storage or processing steps.
What are the rate limits?
You have 60 requests per minute per key. Before your first top-up, the limit is 10 requests per minute. Exceeding the limit causes requests to wait in a queue.
How do I handle errors?
If synthesis fails, the charge is returned in full. Requests are processed sequentially, so under load, you may experience delays. Monitor your response status codes to determine if a request succeeded or failed.
What is the difference between tts-1 and tts-1-hd?
Both models use the same engine and price. They are functionally identical for your integration purposes. Choose either 'tts-1' or 'tts-1-hd' in your model field; the cost remains $0.01 per 1,000 characters.
Can I set the language manually?
Yes. Pass the optional 'language' field with an ISO code. If unset, the system guesses from the script, defaulting to English for Latin text. Explicitly setting it ensures correct pronunciation for non-English Latin-script content.
What happens if I exceed the rate limit?
Requests are processed one after another. Under load, a request may wait in a queue. If you exceed the limit, subsequent requests may be delayed or rejected until the window resets.
Is there a way to stream audio?
No. There is no streaming, WebSocket, or real-time partial audio support. You receive the complete audio file in the response after synthesis finishes. Short phrases take a few seconds; long texts take roughly real-time.
Do I need to install any SDKs?
No. You can use plain HTTP or the official OpenAI SDK by changing the base URL. There is no official custom SDK, mobile SDK, or WordPress plugin required. Any HTTP client works.
Can I use this API in a mobile app?
Yes. You can call the REST API directly from any mobile app. Just include your Bearer token in the Authorization header. Note that storing keys in mobile apps has standard security implications.
First request
curl
curl https://api.zbus.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.mp3Python (openai)
from openai import OpenAI
client = OpenAI(base_url="https://api.zbus.info/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.zbus.info/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 |
Quickstart: Your First Request
Create your ZBus AI account at /get-api-key/ to retrieve your unique API key. This key authenticates your requests to the TTS endpoint and allows you to generate audio files instantly.
Use the provided curl command to send a POST request with your text, selected voice, and preferred format. The API processes your input and returns the audio file directly in the response.
Get your API key
Sign up on the /get-api-key/ page using Google or email. Your dashboard displays your active API key, starting balance, and usage statistics immediately.
- Available via Google or email sign-up
- One active key per account
- Rotate keys anytime via dashboard
Send a curl request
Send a POST request to https://api.zbus.info/v1/audio/speech. Include your API key in the Authorization header and specify the model, input text, voice, and format in the JSON body.
- Endpoint: POST /v1/audio/speech
- Auth: Bearer <your-key>
- Format: JSON
Receive audio file
The API returns the audio file in your chosen format (mp3, wav, opus, flac, or pcm). For texts over 4096 characters, split the input and join the resulting audio files.
OpenAI SDK Compatibility
ZBus AI’s API mirrors the OpenAI speech endpoint structure, allowing you to integrate our text-to-speech engine using your existing OpenAI SDK clients. You only need to adjust the base_url to https://api.zbus.info/v1 while keeping your authentication logic intact.
Use base_url configuration
Point your SDK client to https://api.zbus.info/v1 instead of OpenAI’s servers. Your API key format remains tts-..., ensuring seamless authentication without modifying your authorization headers.
Same request shape
The JSON body accepts model, input, voice, response_format, and speed exactly as defined for the OpenAI speech endpoint. This consistency means your existing code handles our API without complex adaptation layers.
Works with official SDKs
Any official OpenAI SDK (Python, Node.js, etc.) works out of the box. Simply change the base URL and provide your ZBus AI key to start generating audio files for your videos.
- Python SDK
- Node.js SDK
- cURL compatible
- Any HTTP client
Supported Voices
Choose from eight distinct preset voices, each designed for specific content styles. All voices support 23 languages and can speak any of them, making them versatile for diverse creator needs.
Whether you need a warm narrator or a deep documentary tone, our library covers various moods without requiring custom cloning or complex setup.
8 preset voices available
Select from eight high-quality voices derived from licensed LibriTTS-R recordings. Each voice offers a unique tone suitable for different video formats.
All voices are ready to use immediately with no additional configuration required.
- Aria (warm narrator)
- Nova (bright, upbeat)
- Luna (soft storytelling)
- Sage (calm, instructional)
- Felix (energetic promo)
- Milo (friendly conversational)
- Atlas (neutral news)
- Orion (deep documentary)
Female and male options
Our library includes four female and four male voices, ensuring you find the right tone for your audience.
Each voice is optimized for clarity and natural rhythm, ideal for TikTok, Reels, and Shorts content.
Voice ID selection
Specify your chosen voice using its ID in the API request body. The system supports standard voice IDs for easy integration.
You can switch voices instantly between requests to test different tones for your video.
Audio Formats
The API returns audio in five standard formats: MP3, WAV, OPUS, FLAC, and PCM. Choose the format that best fits your video editing workflow or storage needs. Each format is generated on demand with no additional cost or complexity.
Select your preferred format using the response_format field in your request. The service delivers a ready-to-use audio file, ensuring compatibility with most social media platforms and editing software without extra conversion steps.
MP3, WAV, OPUS, FLAC, PCM
All five formats are fully supported for every voice and model. Use MP3 for small file sizes, WAV or FLAC for lossless quality, OPUS for efficient streaming, or PCM for raw audio data.
The choice does not affect the voice characteristics or price. Your audio file arrives exactly as requested, maintaining the quality of the synthetic speech generated by the engine.
- MP3: Compressed, widely compatible
- WAV: Lossless, high fidelity
- OPUS: Efficient compression (for storage/bandwidth)
- FLAC: Lossless compression
- PCM: Raw audio data
Choose format in request
Add the response_format parameter to your JSON body. Valid values are "mp3", "wav", "opus", "flac", or "pcm". The API returns the audio file in the specified format immediately upon completion.
This parameter works with both tts-1 and tts-1-hd models. Ensure your client can handle the returned binary data in the chosen format.
Standard audio quality
Audio quality is consistent across all formats, determined by the model selected rather than the output format. Each generated file carries an inaudible watermark marking it as synthetic speech.
There is no premium tier for higher quality audio. All voices deliver clear, natural-sounding speech suitable for TikTok, Reels, and Shorts content.
Rate Limits
The API allows 60 requests per minute per key for standard usage. This limit applies to the active key on your account, ensuring consistent throughput for your content generation workflow.
60 requests per minute
You can send up to 60 API calls every minute. This rate is sufficient for most creators generating videos for TikTok, Reels, and Shorts without needing complex batching logic.
10 requests before top-up
New accounts with no top-up are limited to 10 requests per minute. Once you add USDT to your balance, the limit increases to 60 requests per minute.
Queue processing under load
Requests are processed sequentially. If you exceed the rate limit or if the system is under heavy load, your request may wait in a queue before synthesis begins.
- Short phrases take a few seconds to generate.
- Longer texts process at roughly real-time speed (25 chars/sec).

Error Handling
When a request fails, the charge is returned in full, so you only pay for successful synthesis. Common issues include hitting rate limits or sending invalid input text that exceeds the 4096-character limit per request. Always validate your input before sending.
If the service is under heavy load, requests may queue behind others. This is normal behavior as requests are processed sequentially. Use the status responses to track your request state and adjust your retry strategy accordingly.
Common error codes
429 Too Many Requests: You have exceeded the 60 requests per minute limit. Wait a moment and retry.
400 Bad Request: The input text is too long or the voice ID is invalid. Check your payload.
503 Service Unavailable: Temporary overload. Your request may be processed later or fail.
Retry logic
Implement exponential backoff for 429 and 503 errors. Start with a short delay and increase it gradually.
Do not retry 400 errors immediately; they indicate a configuration issue that will persist until corrected.
- Wait 1 second, then 2, then 4 seconds between retries.
- Max out retries after 5 attempts to avoid infinite loops.
Status responses
Each response includes a status field indicating success or failure. Failed requests return a full refund to your balance.
Monitor your balance and usage via the dashboard to anticipate limits before they impact your workflow.
Use the status to determine if you should split long texts or wait for load to decrease.
Model Options
The TTS API offers two model identifiers: "tts-1" and "tts-1-hd". Despite the different names, both use the exact same underlying synthesis engine. The distinction is purely for client-side compatibility, allowing you to swap parameters easily if you are migrating from other services or updating your SDK configuration.
tts-1 standard quality
Use "tts-1" as your model identifier for standard generation. This string triggers the default synthesis pipeline and is fully compatible with the official OpenAI SDKs when you adjust the base URL.
tts-1-hd high quality (same engine)
Use "tts-1-hd" for high-quality mode. Although labeled differently, it runs on the identical engine as the standard model. This naming convention ensures your code works seamlessly with libraries expecting the "HD" variant, without any performance or quality difference.
Same engine, same price
Since both models use the same engine, there is no price difference between them. Both are billed at the standard rate of $10 per 1 million characters. You can choose either identifier based on your preference or existing code structure without affecting your billing.
Language Settings
The API automatically detects the language from the input text's script. Latin script defaults to English, but you should set the language field for non-English Latin text to ensure accuracy.
Auto-detection from script
If you omit the language parameter, the system infers it from the characters used. This works well for standard English but may be inaccurate for other Latin-based languages like Dutch or French.
Manual language code
Provide an ISO 639-1 code (e.g., es, jp) to force a specific language. This ensures consistent pronunciation regardless of the input text's appearance.
23 supported languages
Any voice can speak any of the 23 supported languages. Not supported: Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, Bengali, and others.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
Listen to the 8 voices
Each sample below was generated by this API from the same sentence.
23 languages
Any voice speaks any of these languages.
