Frequently asked questions
What is the API endpoint URL?
The endpoint is POST https://api.win9online.com/v1/audio/speech. You send a JSON body containing the text, voice selection, and format preferences. This follows the same request shape as the OpenAI speech endpoint, ensuring compatibility with standard HTTP clients and existing integrations that expect this structure for text-to-speech synthesis.
Which model should I use?
Choose between tts-1 and tts-1-hd. Both models use the same engine and have identical pricing. The choice depends on your quality preference for broadcast-ready audio. Since the engine is the same, performance characteristics like speed and character limits remain consistent.
How do I set the language?
Include the optional language parameter with an ISO code in your request. If omitted, the system guesses the language from the script, defaulting to English for Latin text. For non-English Latin-script text, explicitly setting the language ensures accurate pronunciation. The API supports 23 specific languages for voice synthesis.
Can I use the OpenAI Python SDK?
Yes, the official OpenAI SDKs work by changing the base_url to https://api.win9online.com/v1. Use your site’s API key for authentication. This allows you to leverage existing client code without major modifications. The request shape matches the OpenAI speech endpoint, making integration straightforward for developers already familiar with that ecosystem.
What are the rate limits?
The limit is 60 requests per minute per key. Before your first top-up, the limit is 10 requests per minute. These limits apply to your active API key. If you exceed the limit, requests may queue or fail.
How do I handle long articles?
Split your text into parts of up to 4096 characters per request. Synthesize each part individually, then join the resulting audio files. This approach ensures each request stays within character limits. You can manage this splitting logic in your application code before sending the data to the API endpoint.
What happens if I exceed the limit?
Requests are processed sequentially, so under load a request may wait in a queue. If you exceed the rate limit, subsequent requests might be delayed or rejected until the window resets. Failed requests are free, so you are not charged for audio that does not generate successfully.
Is the API key single-use?
No, the key remains active until you issue a new one. You can generate a new key at any time from your account page. When a new key is issued, the old one stops working immediately. Each account has one active key at a time, ensuring simple and secure access control.
Can I change the voice mid-stream?
No, the API does not support streaming or WebSocket connections. Each request generates a complete audio file for the provided text. To change voices, you must send separate requests for different text segments. You then combine these audio files in your application to create the final broadcast-ready output.
Do I need to install extra packages?
No extra packages are required if you use standard HTTP requests or any HTTP client. If you prefer using the OpenAI Python SDK, you need to install that library and configure the base_url. The API works with any HTTP client capable of sending JSON payloads, so you can use your preferred programming language tools.
How do I authenticate requests?
Include the header Authorization: Bearer tts-... with your unique API key. This key is generated when you sign up or regenerate it from your dashboard. Ensure the key is sent with every request to the API endpoint. The key identifies your account and tracks your usage and billing for the synthesized audio.
What formats are supported?
Supported formats are mp3, wav, opus, flac, and pcm. Specify your preferred format in the response_format field of your request. These options cover most broadcast and digital distribution needs. The API returns the audio file directly in the response, allowing you to save or process it immediately after synthesis.
Is there a webhook for status?
No, there is no webhook support for status updates. The API is synchronous; you receive the audio file or an error response directly in the HTTP response. You must poll or wait for the response to know the status. Request metadata is kept for billing.
How do I regenerate my key?
Regenerate your key from the /get-api-key/ page in your account. A new key is issued immediately, and the old one stops working at once. This ensures secure access management. You can issue a new key anytime; the old key stops working at once.
First request
curl
curl https://api.win9online.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.win9online.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.win9online.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
To begin using NewsAudio, visit the /get-api-key/ page and create an account via Google or email. This step registers your credentials for this specific storefront, allowing you to start generating audio immediately.
Once registered, your dashboard displays your active API key, current balance, and usage statistics. You can test integration directly using the provided curl example before building your production workflow.
Sign up on the dashboard
Navigate to the sign-up page and choose either Google authentication or traditional email/password registration. Google sign-ups include a $0.50 free trial valid for 7 days, while email accounts start with zero credit.
Your account is tied to this specific site (win9online.com). Ensure you are logged into the correct storefront to access your unique credentials and billing history.
View your active key
After logging in, the dashboard prominently displays your active API key. Use this key in the Authorization header (Bearer tts-...) for all API requests. Only one key is active per account at any time.
The interface also shows your current balance, trial status, and character consumption. Use this data to monitor usage against your 60 requests-per-minute limit or your TRC-20 top-ups.
Regenerate key if needed
If you suspect your key has been exposed, you can issue a new key directly from the dashboard. The previous key stops working immediately upon regeneration, ensuring secure access.
Update your integration code promptly after regeneration to avoid authentication errors. Each account maintains a single active key, simplifying security management for editorial teams.
Quickstart with cURL
Use the API key obtained from your dashboard to authenticate requests. Pass it in the Authorization header as a Bearer token. This key identifies your account and tracks usage against your balance.
Send a POST request to the speech endpoint with your text payload. The API processes the input and returns the generated audio file directly in the response body.
Set the Authorization header
Include your key in the request header: Authorization: Bearer tts-your-api-key-here. Ensure the key is active; regeneration invalidates the previous key immediately.
The key must be sent with every request to verify your account and deduct characters from your balance correctly.
- Format:
Authorization: Bearer <key> - Key format starts with
tts- - Regeneration stops the old key instantly
Send a POST request to /v1/audio/speech
Send a JSON body containing the model, input, and voice. The endpoint is https://api.win9online.com/v1/audio/speech. Limit text to 4096 characters per request.
Use tts-1 or tts-1-hd for the model. Specify the voice preset and desired audio format like mp3 or wav.
- Endpoint:
POST /v1/audio/speech - Max input: 4096 characters
- Models:
tts-1,tts-1-hd
Receive audio file in response
The response contains the audio data directly. Save the output to a file with the appropriate extension (e.g., .mp3, .wav).
If synthesis fails, the request is free and the charge is returned in full.
- Response contains raw audio data
- Supported formats: mp3, wav, opus, flac, pcm
- Failed requests are free

Using the OpenAI SDK
Integrate the official OpenAI SDKs into your newsroom’s backend by simply overriding the default API base URL. Point the client to https://api.win9online.com/v1 while retaining your existing library structure. This allows editorial teams to use familiar code patterns without learning a new client wrapper.
Change base_url to api.win9online.com
Update your SDK configuration to target https://api.win9online.com/v1. Ensure you pass your NewsAudio API key (starting with tts-...) in the Authorization: Bearer header. This single configuration change directs all synthesis requests to our cloud infrastructure while preserving standard OpenAI compatibility.
Use tts-1 or tts-1-hd models
Select either tts-1 or tts-1-hd as your model parameter. Both utilize the same underlying engine and pricing structure. Choose tts-1 or tts-1-hd; they use the same engine and have identical pricing.
Maintain same request shape
The JSON body structure remains identical to the OpenAI standard. Pass your text via the input field, select a voice from the eight presets, and specify the desired response_format (mp3, wav, opus, flac, or pcm). Speed adjustments between 0.25 and 4.0 are also supported within this standard schema.
- Pass text via the
inputfield - Select a voice from eight presets
- Specify
response_format(mp3, wav, opus, flac, pcm) - Adjust speed between 0.25 and 4.0
Request Parameters
The API accepts plain text input, with a hard limit of 4,096 characters per request. Spaces and punctuation count toward this total. For longer articles, split the text into chunks, generate audio for each, and join the files client-side.
Select one of eight voice presets to match your editorial tone. Each voice is based on licensed recordings from LibriTTS-R and LibriVox. You can also specify the response format, choosing from mp3, wav, opus, flac, or pcm depending on your storage and playback needs.
input: text up to 4096 chars
Send your script in the input field. The engine processes up to 4,096 characters per request. If your article exceeds this limit, split it into smaller parts. Generate audio for each chunk and concatenate the resulting files to reconstruct the full piece.
voice: select from 8 presets
Choose a voice that fits your broadcast style. Options include female voices like Aria (warm narrator) and Nova (upbeat), or male voices like Felix (energetic promo) and Atlas (neutral news). Each voice is based on licensed recordings from LibriTTS-R and LibriVox, ensuring consistency across your library.
- 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: mp3, wav, etc.
Specify the output format using the response_format parameter. Supported formats include mp3, wav, opus, flac, and pcm. Choose the format that best suits your editorial workflow and storage requirements. The API returns the audio file directly in the response body.
- mp3
- wav
- opus
- flac
- pcm
Rate Limits and Queues
Each API key allows up to 60 requests per minute. This limit applies across all concurrent connections using that key. If you exceed this threshold, the API returns a 429 Too Many Requests error.
60 requests per minute per key
The rate limit is strict and enforced per key. You can monitor your usage in the dashboard to ensure you stay within the 60 RPM threshold. If you need higher throughput, consider distributing requests across multiple keys.
- Limit: 60 requests/minute/key
- Error code: 429
10 requests before first top-up
New accounts start with a reduced limit of 10 requests per minute. This allows you to test the integration before committing to a top-up. Once you add funds via USDT (TRC-20), the limit increases to the standard 60 RPM.
Queue processing under load
Requests are processed sequentially within your allocated rate limit. Under heavy load, additional requests may wait in a queue before processing begins. This ensures stable performance but may increase latency for bursty traffic patterns.

Handling Long Texts
The API enforces a strict 4096-character limit per request. To convert full-length articles or long scripts, you must split the input text into smaller segments that fit within this constraint before sending them to the endpoint.
Split text into 4096 char chunks
Divide your source text into blocks of up to 4096 characters. Spaces and punctuation count toward this limit. Ensure you do not exceed this boundary, as requests longer than the limit will fail.
Generate audio for each part
Send each chunk as a separate POST request to /v1/audio/speech. The API processes requests sequentially, so longer texts may experience queue delays. Each successful request returns a distinct audio file.
Join audio files client-side
Receive the individual audio segments in your application. Concatenate these files in the correct order to reconstruct the complete broadcast-ready audio file. This ensures consistent voice and tone throughout the entire script.
Error Handling
The API returns standard HTTP status codes to indicate request outcomes. A 200 OK response contains the generated audio file in the specified format. Errors such as 400 (bad request) or 401 (unauthorized) provide details in the JSON body to help debug invalid inputs or authentication issues.
Check status codes
Always inspect the HTTP status code before parsing the response body. A 2xx code indicates success, while 4xx or 5xx codes signal client or server errors. The response body for errors includes a message explaining the issue, such as an invalid voice ID or exceeded character limit.
Handle 429 rate limit errors
When you exceed the limit of 60 requests per minute, the API returns a 429 Too Many Requests status. Your key is limited to 10 requests per minute before the first top-up. Wait for the rate limit window to reset before sending more requests.
- Return code: 429
- Cause: Exceeded 60 requests/minute
- Action: Wait and retry
Retry failed requests
Implement exponential backoff for transient errors like 503 Service Unavailable or 429 Rate Limit. Since requests are processed sequentially, brief delays often resolve queue congestion. Failed synthesis requests are free, so retrying after a short pause does not incur additional charges if the error persists.

Voice and Language Options
Select one of eight distinct voice presets to match your editorial tone. Options range from warm narrators like Aria to neutral news readers like Atlas, ensuring consistent audio branding across your broadcasts.
Set voice parameter explicitly
Choose from eight licensed presets including Aria, Nova, Luna, Sage, Felix, Milo, Atlas, and Orion. Each voice is based on real recordings, providing a professional, consistent sound for your articles.
- Aria: warm, clear narrator
- Atlas: neutral, news
- Felix: energetic, promo
Optional ISO language code
Send an ISO language code (e.g., 'de', 'fr', 'es') to specify the language explicitly. This ensures accurate pronunciation and intonation, especially when your script contains mixed languages or ambiguous phrasing.
- Supports 23 languages including Arabic, Chinese, German, Japanese, and Spanish.
- Any voice can speak any supported language.
Auto-detect for Latin script
If you omit the language parameter, the API guesses the language from the script. Text in Latin script defaults to English. For non-English Latin text, always set the language to avoid errors.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. The news publication team collaborates across large divisions to ensure accurate and timely reporting.
23 languages
Any voice speaks any of these languages.
