API Reference: Voices, Pricing & Integration

By VocalLedger editorial team · Updated 2026-11-18

This documentation provides a practical reference for product teams in banking and fintech who need to integrate reliable voice notifications into their alert systems. You will find clear examples for generating speech via a simple HTTP API, covering authentication, available voices, and transparent pricing structures.

Learn how to send text and receive audio files using standard OpenAI SDKs or direct cURL commands. The guide details rate limits, error handling, and data handling specifics, ensuring you can implement fraud warnings and updates with predictable performance and cost.

API Reference: Voices, Pricing & Integration

Frequently asked questions

How do I get an API key?

Sign up on the /get-api-key/ page using Google or email and password. Accounts are specific to this site. The dashboard displays your active API key, balance, and usage statistics immediately after registration. You can issue a new key at any time, which instantly invalidates the previous one.

What is the base URL for the API?

The API base URL is https://api.playga26.com/v1. Use the POST endpoint https://api.playga26.com/v1/audio/speech to generate audio. Other endpoints include GET /v1/voices for voice details and GET /v1/models for model information. All requests must be sent to these specific paths.

Can I use the OpenAI SDK?

Yes, the official OpenAI SDKs work by changing the base_url to https://api.playga26.com/v1. Use the site's API key for authentication. The request shape matches the OpenAI speech endpoint exactly. This allows you to integrate our TTS API using libraries you likely already know.

What are the rate limits?

You are limited to 60 requests per minute per key. Before your first top-up, the limit is 10 requests per minute. Requests are processed sequentially, so under load, a request may wait in a queue. Ensure your application handles these limits to avoid delays.

How do I handle errors?

Failed requests are free, and the charge is returned in full if synthesis fails. Monitor your response status codes to detect errors. If a request fails, no characters are deducted from your balance. You can retry the request or split your text if it exceeds limits.

What models are available?

Two models are available: tts-1 and tts-1-hd. Both use the same engine and have identical pricing. You can specify either model in your request body. The choice depends on your quality preference, but the underlying synthesis technology remains consistent across both options.

How do I list available voices?

Send a GET request to /v1/voices. This endpoint returns details on available voices, languages, formats, and pricing. You will see eight presets, including female voices like Aria and Nova, and male voices like Felix and Milo. Use this data to select the voice for your content.

Is authentication required?

Yes, authentication is required for all API requests. Include your API key in the Authorization header as 'Bearer tts-...'. Each account has one active key. If you issue a new key, the old one stops working immediately. Ensure your key is valid before sending requests.

What happens if I exceed the rate limit?

Requests are processed sequentially, so exceeding the limit may cause your request to wait in a queue. The standard limit is 60 requests per minute, or 10 before the first top-up. Your application should handle delays gracefully. Failed requests do not incur charges.

Can I use multiple keys?

Each account has one active key. You can issue a new key at any time, which instantly invalidates the previous one. This allows you to rotate keys for security or manage different usage tiers. Only one key is active per account at any given time.

How do I specify a language?

Include the optional 'language' field with an ISO code in your request body. If omitted, the system guesses the language from the script, defaulting to English for Latin text. For non-English Latin-script text, explicitly set the language to ensure accurate synthesis and pronunciation.

What is the maximum input length?

The maximum input length is 4096 characters per request. For longer texts, split them into parts of up to 4096 characters and join the resulting audio files. Spaces and punctuation count toward the character limit. This ensures consistent processing and predictable generation times for your content.

Are there any SDKs for mobile?

No official mobile SDK exists. You can use the standard OpenAI SDK or plain HTTP requests to integrate with mobile applications. Since the API follows standard REST patterns, you can easily implement it using any HTTP library available for your mobile platform, such as iOS or Android.

Do I need to install any libraries?

No specific library is required. You can use plain HTTP requests or the official OpenAI SDK by adjusting the base URL. This allows you to use existing tools in your tech stack. No proprietary libraries or plugins need to be installed to start generating audio files.

First request

curl

curl https://api.playga26.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.mp3

Python (openai)

from openai import OpenAI

client = OpenAI(base_url="https://api.playga26.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.playga26.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
ariafemale
novafemale
lunafemale
sagefemale
felixmale
milomale
atlasmale
orionmale
language
arالعربية
zh中文
daDansk
nlNederlands
enEnglish
fiSuomi
frFrançais
deDeutsch
elΕλληνικά
heעברית
hiहिन्दी
itItaliano
ja日本語
ko한국어
msBahasa Melayu
noNorsk
plPolski
ptPortuguês
ruРусский
esEspañol
swKiswahili
svSvenska
trTürkçe

Authentication

Sound wave illustration for: Authentication

Access your unique API key by signing in at /get-api-key/. The dashboard displays your active key alongside your current balance, character usage, and a ready-to-use cURL example.

Get your API key from the dashboard

Sign up using Google or email/password to receive your key. The dashboard provides immediate access to your credentials and usage statistics, ensuring you can start generating audio notifications for your banking or fintech applications without delay.

Use Bearer token in Authorization header

Include your key in the HTTP header as Authorization: Bearer tts-.... This standard authentication method allows you to use the official OpenAI SDKs by simply pointing them to our API base URL.

  • Format: Authorization: Bearer tts-<your-key>
  • Supported by official OpenAI SDKs

One active key per account

Each account maintains a single active API key. If you need to revoke access, you can issue a new key at any time; the previous key becomes invalid immediately.

  • One active key per account
  • Instant invalidation of old keys

First Request with cURL

Sound wave illustration for: First Request with cURL

Send a POST request to the API endpoint with your Bearer token. The payload requires the model, input text, and voice preset. For banking alerts, use clear tones like atlas for news-style updates or milo for conversational notifications.

Include input with your alert text (max 4096 characters). Choose response_format such as mp3 or wav based on your playback system’s capabilities. The API returns the audio file directly in the response body for immediate use.

Basic syntax for POST request

Use cURL to send your first synthesis request. Ensure your API key is correctly formatted in the Authorization header.

The request body must be valid JSON containing the required parameters for the TTS engine to process your text.

  • Endpoint: https://api.playga26.com/v1/audio/speech
  • Method: POST
  • Header: Authorization: Bearer tts-...

Required fields: model, input, voice

Specify a voice preset like atlas for news-style updates. The input field carries your notification message, and voice selects the speaker tone appropriate for your audience, such as milo for conversational alerts or atlas for formal banking updates.

Choose response_format such as mp3 or wav. The model field accepts tts-1 or tts-1-hd, which use the same engine and price.

Example response with audio file

Upon success, the API returns the audio file directly in the response body. You can save this file to your servers or stream it to your users. Failed requests return a standard error code without charging your balance.

First Request with cURL

OpenAI SDK Integration

Sound wave illustration for: OpenAI SDK Integration

Integrate VocalLedger’s TTS API into your fintech notification workflows using the official OpenAI SDKs. This approach allows product teams to maintain familiar code structures while switching the underlying voice engine. You simply direct the SDK to our API base URL, enabling seamless text-to-speech generation for fraud alerts and account updates.

The integration requires minimal configuration changes. By pointing the official OpenAI SDK to our base URL, you leverage familiar methods like client.audio.speech.create(). This ensures consistent behavior with standard parameters, allowing your team to generate audio files efficiently without learning a new client library or handling complex HTTP requests manually.

Install the OpenAI Python/Node SDK

Install the standard OpenAI package for your preferred language. For Python, use pip install openai. For Node.js, run npm install openai. This provides the necessary client classes to handle API interactions without requiring custom HTTP wrappers or additional dependencies.

The SDK manages request serialization and response parsing automatically. This reduces boilerplate code in your notification service, ensuring that text payloads are correctly formatted and audio responses are handled efficiently within your existing fintech infrastructure.

  • Python: pip install openai
  • Node.js: npm install openai

Set base_url to our endpoint

Configure the SDK to use our API domain instead of OpenAI’s servers. Set the base_url to https://api.playga26.com/v1. This directs all subsequent speech generation requests to VocalLedger’s infrastructure while maintaining the standard SDK interface.

Ensure your API key is passed via the standard OPENAI_API_KEY environment variable or explicitly in the client configuration. The SDK will automatically attach this key as a Bearer token in the Authorization header for each request.

  • base_url: https://api.playga26.com/v1
  • Auth: Bearer token via standard header

Use standard speech parameters

Pass text, voice, and model parameters exactly as you would for OpenAI. Use model set to tts-1 or tts-1-hd. Select a voice from the 8 available presets, such as atlas for neutral news-style alerts or nova for upbeat updates. Specify the input text up to 4096 characters.

Set the response_format to your preferred audio type, such as mp3 or wav. The SDK returns the audio file directly in the response, which your application can then stream or save. This consistency allows you to swap voice engines without rewriting your notification logic.

  • model: tts-1 or tts-1-hd
  • voice: e.g., atlas, nova
  • input: max 4096 characters
  • response_format: mp3, wav, opus, flac, pcm

Endpoint: /v1/audio/speech

Sound wave illustration for: Endpoint: /v1/audio/speech

The /v1/audio/speech endpoint accepts POST requests to synthesize text into audio files. Send your API key in the Authorization header and include the text payload in the JSON body. This is the core endpoint for generating speech notifications, alerts, or updates for your fintech applications.

POST method for speech generation

Send a POST request to https://api.playga26.com/v1/audio/speech. Include your Bearer token in the Authorization header. The request body must contain the text to convert, the voice preset, and the model type. Ensure your text does not exceed 4096 characters per request.

JSON body with text and settings

Use the standard OpenAI-compatible JSON structure. Required fields are model ("tts-1" or "tts-1-hd"), input (your text), and voice (one of the eight presets). Optional parameters include response_format, speed, and language. If language is omitted, the system guesses from the script, defaulting to English for Latin text.

  • model: "tts-1" or "tts-1-hd"
  • input: string up to 4096 characters
  • voice: e.g., "aria", "nova", "atlas"
  • response_format: mp3, wav, opus, flac, pcm
  • speed: 0.25–4.0
  • language: optional ISO code

Audio file returned in response

The response returns the generated audio file directly. Failed requests are free, with charges returned in full. For longer texts, split into parts under 4096 characters and join the resulting audio files. Each request is processed sequentially, so queue delays may occur under load.

Endpoint: /v1/voices

Sound wave illustration for: Endpoint: /v1/voices

Retrieve the full catalog of available voices, supported languages, and output formats by sending a GET request to this endpoint. The response provides the metadata necessary to construct valid speech synthesis requests, ensuring your integration uses only compatible parameters.

GET method to list available voices

Calling this endpoint returns a JSON object listing all eight voice presets. Each entry includes the voice ID, associated language, and available response formats. This data is essential for dynamic UI generation, allowing your application to present voice options directly to end-users without hardcoding values.

  • Returns voice ID, name, and language code.
  • Lists supported response formats (mp3, wav, opus, flac, pcm).
  • Includes base price per character for transparency.

Details on language and format support

The API supports 23 languages, including English, Spanish, German, French, Arabic, Japanese, and Chinese. Any voice can speak any of these languages; if the language is not specified, the system guesses it from the script, defaulting to English for Latin text. For non-English Latin-script text, explicitly set the language parameter to avoid misidentification.

  • Supports 23 languages including Swahili, Turkish, and Greek.
  • Format options: mp3, wav, opus, flac, pcm.
  • Language detection is automatic but can be overridden.

Voice descriptions for selection

Select voices based on their tone: Aria for warm narration, Nova for bright updates, Luna for soft storytelling, Sage for calm instructions, Felix for energetic promos, Milo for friendly conversation, Atlas for neutral news, and Orion for deep documentaries. These presets are based on licensed recordings from LibriTTS-R and LibriVox public-domain audiobook readings, ensuring clear, synthetic speech with an inaudible watermark.

Endpoint: /v1/voices

Endpoint: /v1/models

Sound wave illustration for: Endpoint: /v1/models

The /v1/models endpoint provides a simple list of available synthesis engines, allowing your application to dynamically detect supported versions. This ensures your integration remains compatible as we update our infrastructure without breaking existing requests.

GET method to list models

Send a GET request to retrieve the current list of active models. This lightweight call helps your system verify available options before initiating speech generation, ensuring you always target a valid engine version.

  • Returns a JSON array of model IDs
  • Includes model name and version details
  • No authentication required for basic discovery

tts-1 and tts-1-hd options

The endpoint lists two models: tts-1 and tts-1-hd. Both utilize the same underlying synthesis engine, delivering consistent voice quality and pronunciation behavior across all eight presets like Aria, Nova, and Atlas.

  • tts-1: Standard latency performance
  • tts-1-hd: High-definition audio output
  • Both support 8 distinct voice presets

Same engine and price for both

Since both models share the same engine, there is no functional difference in voice character or language support. Pricing is identical at $10 per 1 million characters, allowing you to choose based on audio format needs without cost variance.

  • Identical character-based pricing
  • No performance difference in voice output
  • Easy switching between models

Rate Limits

Sound wave illustration for: Rate Limits

Text-to-speech requests are processed sequentially, meaning under load a request may wait in a queue. Long texts are split into 4096-character parts, and each part counts toward your limit. This ensures reliable generation for fraud alerts and notifications without overwhelming the service.

60 requests per minute per key

Once your account is active with a top-up, you can send up to 60 requests per minute. This throughput is sufficient for most notification workflows, allowing product teams to batch-process alerts efficiently without hitting hard caps during standard usage peaks.

10 requests per minute initially

New accounts start with a limit of 10 requests per minute. This lower baseline prevents accidental bursts from consuming trial credit too quickly. It is ideal for initial integration testing and verifying that your text formatting and voice selection work correctly.

Limit increases after first top-up

The higher 60 requests per minute limit activates automatically after your first top-up. You do not need to manually adjust settings or contact support. Simply add USDT (TRC-20) to your balance, and the increased throughput applies immediately to your API key.

Rate Limits

Error Handling

Sound wave illustration for: Error Handling

When integrating the VocalLedger API, robust error handling ensures your banking or fintech applications remain resilient during high-volume notification bursts. Understanding HTTP status codes and synthesis failures allows product teams to distinguish between transient network issues and service-side constraints, ensuring accurate billing and reliable user alerts.

Failed synthesis requests are handled gracefully: if the engine cannot produce audio, the charge is returned in full. This guarantees that your budget remains aligned with actual successful generations, a critical factor for cost-sensitive notification pipelines in financial services.

Common HTTP error codes

The API returns standard HTTP codes to indicate request status. A 4xx error typically indicates an issue with your authentication key or request format, such as an invalid voice ID or exceeding the 4096-character limit. A 5xx error suggests a temporary server issue, which you should handle with exponential backoff.

  • 400: Bad Request (invalid parameters or format)
  • 401: Unauthorized (invalid or expired API key)
  • 429: Too Many Requests (rate limit exceeded)
  • 500-503: Server errors (transient, retry with backoff)

Handling failed synthesis requests

If speech synthesis fails due to engine errors, the API returns a non-200 status code along with an error object detailing the issue. Crucially, no charge is applied for these failed attempts, and any prepaid balance remains intact. This ensures financial accuracy for your notification system.

You should log these failures to monitor engine health. Since requests are processed sequentially, a single bottleneck can delay subsequent requests, but failed ones do not consume your per-minute quota beyond the initial attempt.

Retry logic for queue delays

During peak usage, requests may queue behind longer synthesis tasks, causing slight delays. Implement exponential backoff with jitter for 429 or 5xx responses. Avoid immediate retries, as the queue may still be processing prior requests.

For time-sensitive fraud alerts, consider sending the request well in advance of the user-facing deadline. If a retry is needed, ensure your application logic accounts for the potential latency without duplicating the notification to the end user.

  • Use exponential backoff for transient errors
  • Add jitter to prevent thundering herd effects
  • Check for 429 status to respect rate limits

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. Secure your funds instantly with our new real-time fraud alerts and smart transaction monitoring features.

Avatar of the voice aria
ariafemale
Avatar of the voice nova
novafemale
Avatar of the voice luna
lunafemale
Avatar of the voice sage
sagefemale
Avatar of the voice felix
felixmale
Avatar of the voice milo
milomale
Avatar of the voice atlas
atlasmale
Avatar of the voice orion
orionmale

23 languages

Any voice speaks any of these languages.

Sources and further reading

  1. Wikipedia
  2. OpenAI API reference: create speech
  3. Tether (USDT)
  4. ISO 639 language codes
  5. TRON developer hub: TRC-20
  6. IETF RFC 7845: Ogg Opus
  7. MDN: Web Speech API
  8. Python documentation
  9. LibriTTS-R paper (arXiv 2305.18802)
  10. LibriTTS-R corpus (OpenSLR 141)
  11. W3C WAI: making audio and video accessible
  12. Creative Commons Attribution 4.0

More on this site