What is the Text-to-Speech API and how does it work?

By DubSync editorial team · Updated 2026-11-18

DubSync API is a cloud-based text-to-speech service designed for freelancers and localization agencies who need reliable voiceovers for dubbing projects. You send text via a standard JSON endpoint and receive an audio file after synthesis completes, fitting seamlessly into your existing workflow without complex integrations.

This page explains how our API works, from the 8 licensed voice presets to the 23 supported languages. You will learn about our per-character pricing, technical limits like the 4096-character cap, and why our OpenAI SDK compatibility makes integration straightforward for your next localization project.

What is the Text-to-Speech API and how does it work?

Frequently asked questions

What does TTS stand for?

TTS stands for Text-to-Speech. Our API converts written text into audio files via a simple HTTP POST request. You send the text, and we return the synthesized audio. It is designed to integrate easily into your existing localization and dubbing workflows without requiring complex infrastructure.

Is the TTS API free to use?

Yes, with a free trial. New Google sign-ups receive $0.50 credit (about 50,000 characters) valid for 7 days. After that, usage is pay-as-you-go at $0.01 per 1,000 characters. Failed requests are free, and there are no monthly plans or minimum spend requirements.

How many languages does it support?

The API supports 23 languages, including English, Spanish, German, French, Japanese, Korean, Chinese, Arabic, Hindi, and Portuguese. Any voice can speak any of these languages. If the language is not specified, it is guessed from the script, defaulting to English for Latin text.

Can I use the voices for commercial projects?

Yes. You can use the output commercially. The voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and LibriVox public-domain audiobook readings. You can use the generated audio for commercial dubbing, videos, or applications. Each file includes an inaudible watermark to mark it as synthetic speech.

Does it support SSML?

No, the API does not support SSML. You send plain text input, and the system handles the synthesis. There are no emotion tags or custom pronunciation lexicons available. For precise control, you can adjust the speed between 0.25 and 4.0.

Can I clone my own voice?

No, we do not offer custom voice cloning. You cannot upload your own voice or design new voices. You must choose from the 8 available presets, such as Aria, Nova, Felix, or Atlas, which are based on licensed recordings from public-domain sources.

How long does generation take?

Generation is roughly real-time, processing about 25 characters of text per second. Short phrases take a few seconds, while longer texts are split into parts and joined. Requests are processed sequentially, so under load, a request may wait in a queue.

Is there a character limit?

Yes, each request supports up to 4096 characters. For longer texts, you should split the input into smaller parts and join the resulting audio files. Spaces and punctuation count toward the character limit and the billing calculation.

Can I use it for dubbing?

Yes, it is built for freelancers and agencies doing dubbing. You can send text and receive audio files in formats like mp3, wav, opus, flac, or pcm. The API fits directly into your localization workflow, allowing you to generate voiceovers efficiently.

How do I pay for the service?

Payments are made via USDT on the TRON network (TRC-20). The minimum top-up is $5, and credits are added automatically within minutes. We do not accept cards, PayPal, or subscriptions. Your balance does not expire, and there are no monthly fees.

Is there a free trial?

Yes, signing in with Google gives you $0.50 in credit, valid for 7 days. This covers about 50,000 characters. Accounts created with email and password start without trial credit. You can top up with USDT whenever you need more capacity.

Can I use it on mobile devices?

The API is cloud-based and accessed via HTTP requests, so it works on any device with internet access. There is no mobile SDK.

Does it work with OpenAI SDK?

Yes, the request shape matches the OpenAI speech endpoint. You can use the official OpenAI SDK by changing the base_url to https://api.playtk71.com/v1 and using your site’s key. This allows for easy integration into existing codebases without rewriting logic.

Can I get a refund for failed requests?

Yes, failed requests are free. If synthesis fails, the charge is returned in full. We keep request metadata for billing.

Core functionality

Sound wave illustration for: Core functionality

Send plain text to a single API endpoint and receive a synthesized audio file. The service converts your script into natural-sounding speech using modern cloud infrastructure, handling the heavy lifting of synthesis so you can focus on your localization workflow.

Text to audio conversion

Input your text and receive an audio file in seconds. The engine processes up to 4096 characters per request, returning MP3, WAV, OPUS, FLAC, or PCM formats. Failed requests are free, ensuring you only pay for successful synthesis.

  • JSON body input with standard fields
  • Supports multiple audio output formats
  • Automatic character counting for billing

Cloud-based processing

All synthesis happens on our servers, requiring no local GPU or software installation. Requests are processed sequentially, meaning longer texts may queue under load and take longer to complete. You simply send a POST request and wait for the result.

API-driven access

Access is fully programmatic via a standard REST API compatible with the OpenAI SDK. Use a simple cURL command or integrate directly into your Python or Node.js scripts. Authenticate with a Bearer token and manage voices, models, and usage via clear JSON responses.

  • OpenAI SDK compatible
  • Standard JSON endpoints
  • Bearer token authentication

Voice technology

Sound wave illustration for: Voice technology

Voices are based on licensed recordings: LibriTTS-R (CC BY 4.0), from LibriVox public-domain audiobook readings., ensuring a solid foundation for natural-sounding dubbing.

Licensed LibriTTS-R base

Our technology rests on licensed LibriTTS-R data, sourced from LibriVox public-domain audiobook readings. This foundation provides a consistent, high-quality baseline for all generated speech, tailored for localization workflows.

8 distinct voice presets

Choose from 8 specific presets designed for different tonal needs. Select from female voices like Aria, Nova, Luna, and Sage, or male options including Felix, Milo, Atlas, and Orion. Each voice offers a distinct character suitable for various dubbing styles.

  • 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

Inaudible synthetic watermark

Every audio file generated through our API includes an inaudible watermark. This marker clearly identifies the content as synthetic speech, which is useful for tracking and verifying voiceover origins in your production pipeline.

Supported languages

Sound wave illustration for: Supported languages

The API supports 23 languages, covering major markets for localization. Any voice can speak any of these languages, giving you flexibility in casting.

We support Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. Languages like Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, and Bengali are not supported.

23 languages available

Our engine handles 23 specific languages for dubbing and voiceover projects. This list covers the majority of global localization needs.

Supported languages include Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. Unlisted languages like Vietnamese or Thai are not supported.

  • 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 of language

If you do not specify a language, the system guesses it from the text script. Latin script text defaults to English.

This simplifies your workflow for English content, but relies on script detection.

Manual language override

For non-English text using Latin script, always set the language parameter. This ensures correct pronunciation and accent.

Pass the ISO code in the request body to override automatic detection.

Audio formats

Sound wave illustration for: Audio formats

The API supports multiple audio formats to fit different delivery needs. You can choose MP3 for universal compatibility, WAV for uncompressed quality, OPUS for efficient compression, or FLAC for lossless archival. PCM is available for raw audio data.

Format selection is handled directly in the request body via the response_format field. This allows you to optimize file size and quality for each specific dubbing or voiceover task without needing post-processing.

MP3, WAV, OPUS, FLAC

MP3 offers a good balance of quality and file size, making it ideal for web playback and standard video dubbing. WAV provides uncompressed, high-fidelity audio suitable for professional editing workflows. OPUS delivers excellent compression for low-bandwidth scenarios, while FLAC ensures lossless preservation for archival purposes.

All these formats are generated directly from the TTS engine, ensuring consistent quality regardless of the output type.

PCM for raw data

PCM (Pulse Code Modulation) returns raw audio samples without any container overhead. This is useful if you are building a custom pipeline or need to process the audio data immediately after generation.

Use this format when you prefer to handle encoding and containerization yourself, saving server-side processing time for container creation.

Format selection per request

The format is specified in the JSON request body using the response_format field. You can change the format for each individual request, allowing flexibility in your workflow.

This per-request control ensures you only pay for the format you actually need, avoiding unnecessary storage or bandwidth costs.

Integration capabilities

Sound wave illustration for: Integration capabilities

The API mirrors the OpenAI speech endpoint structure, allowing you to use the official OpenAI SDKs with a simple base_url change to https://api.playtk71.com/v1. This allows you to use the official OpenAI SDKs by changing the base_url, enabling quick integration into your existing localization scripts.

Standard JSON endpoints handle all interactions, making it easy to automate voice generation within your dubbing pipeline. You can send text and receive audio files directly, streamlining your workflow from script to final audio asset without manual intervention.

OpenAI SDK compatibility

The API mirrors the OpenAI speech endpoint structure, allowing you to use the official OpenAI SDKs with a simple base_url change to https://api.playtk71.com/v1. This eliminates the need for custom wrappers or learning a new request schema, enabling quick integration into your existing localization scripts.

  • Use official OpenAI SDKs
  • Change base_url only
  • No custom wrappers needed

Standard JSON endpoints

Standard JSON endpoints handle all interactions, making it easy to automate voice generation within your dubbing pipeline. You can send text and receive audio files directly, streamlining your workflow from script to final audio asset without manual intervention.

Easy curl integration

For quick testing or lightweight scripts, a single curl command can generate speech. Simply provide your API key, select a model and voice, and pass your text to receive the audio file in your desired format.

This simplicity allows freelancers to prototype dubbing projects rapidly before building full-scale automation in Python or JavaScript.

  • Quick prototyping
  • Lightweight scripts
  • Direct file retrieval

Pricing model

Sound wave illustration for: Pricing model

Our pricing is straightforward: you pay only for what you process. The rate is $10 per 1 million characters, which equals $0.01 per 1,000 characters. Spaces and punctuation count toward the total. If a synthesis fails, you are not charged. Failed requests receive a full refund of the character cost.

There are no monthly subscriptions, no seats to manage, and no minimum spend requirements. You can top up your balance with USDT on the TRON network (TRC-20) starting at just $5. Credits appear automatically within minutes, and your balance never expires. We do not offer invoices with VAT or credit card payments.

Per-character billing

You pay $10 per 1 million characters processed. This includes spaces and punctuation. Failed requests are free, with the charge returned in full.

No monthly fees

There are no monthly plans, no per-seat costs, and no minimum spend. We do not offer subscriptions, invoices with VAT, or credit card billing. Top-ups via USDT on TRON (TRC-20) start at $5 and credit automatically within minutes.

Free trial available

Sign in with Google to receive $0.50 in trial credit (about 50,000 characters), valid for 7 days. Accounts created with email and password start without trial credit.

Use cases for freelancers

Sound wave illustration for: Use cases for freelancers

Freelancers in localization need reliable, fast voice generation that fits seamlessly into their daily workflow. Our API provides the cloud TTS engine designed for agencies and solo developers who need high-quality audio without managing infrastructure. You send text, we return the audio file, allowing you to focus on the creative and technical aspects of your project.

Dubbing short videos

Generate voiceovers for short-form content quickly. With support for 23 languages, you can match the source video's language or adapt it for new markets. Use the speed parameter to adjust playback speed, ensuring natural pacing for social media or marketing content.

  • Support for 23 languages including English, Spanish, and Japanese
  • Adjustable speed from 0.25x to 4.0x for precise sync
  • Instant audio delivery via standard API response

Creating audiobooks

Produce narration for indie books or podcasts. Split long texts into chunks under 4,096 characters and join the resulting audio files. Choose from eight distinct voice presets, such as Luna for soft storytelling or Atlas for neutral news, to match your content's tone.

  • 8 distinct voice presets for varied narration styles
  • Split and join method for texts exceeding 4,096 characters
  • High-quality formats like WAV and FLAC available

Generating voiceovers

Create voiceovers for e-learning modules, explainer videos, or game assets. The API accepts plain text and returns audio in MP3, WAV, OPUS, or FLAC. This flexibility lets you choose the best format for your specific delivery channel without extra conversion steps.

  • Multiple output formats: MP3, WAV, OPUS, FLAC, PCM
  • Simple JSON body request compatible with OpenAI SDKs
  • No need for manual audio editing or conversion
Use cases for freelancers

Data privacy

Sound wave illustration for: Data privacy

Your input text is used solely to generate the audio file and is never added to our training data. This ensures your unique scripts remain yours, whether you are localizing a short video or producing an audiobook for a client.

Once the audio response is delivered, the generated file is not stored on our servers. This minimizes storage risks and keeps your content secure during the generation process.

No training on input text ['The text you send via the API is processed exclusively for synthesis. It is not used for training., so your creative assets remain private and distinct from our dataset.', {'h3': 'Audio not stored after response', 'paras': ['Generated audio files are returned directly in the response and are not retained by our infrastructure after delivery. This ensures your final voiceovers are not stored unnecessarily, reducing potential privacy concerns for sensitive dubbing projects.', {'h3': 'Metadata kept for billing', 'paras': ['We retain request metadata—such as timestamp, character count, and status—strictly for accurate billing and usage tracking. This data helps us process your charges correctly without storing the actual audio content or input text.', 'bullets']}, 'bullets']}, {'h3': 'Support and contact', 'paras': ['For any questions regarding data handling or usage, our support team is available via email. Documentation is also available for detailed technical references.', 'bullets']}, {'image': 'A clean, modern server rack in a dimly lit data center with soft blue LED lights, representing secure cloud infrastructure for text-to-speech processing, no people, no logos'}]

Technical limitations

Sound wave illustration for: Technical limitations

The API imposes a strict 4096-character limit per request. For longer texts, you must split the content into chunks and join the resulting audio files manually or via script. This ensures consistent processing quality but requires basic logic in your workflow.

Generation is not real-time. Short phrases take a few seconds, and longer texts process at roughly real-time speed (25 characters per second). Requests are queued, so under load, you may experience slight delays.

4096 character limit per request

Each API call accepts up to 4096 characters, including spaces and punctuation. To generate audio for longer scripts, split your text into parts within this limit and concatenate the output files.

This limit applies to the input field in the JSON body. Ensure your code handles the splitting logic before sending requests to the POST /v1/audio/speech endpoint.

  • Max 4096 characters per request.
  • Includes all spaces and punctuation.
  • Requires manual or scripted chunking for long texts.

No streaming support

The API does not support WebSocket or real-time partial audio streaming. You receive the complete audio file only after synthesis finishes.

This means you cannot play audio word-by-word as it generates. For dubbing workflows, this is ideal for precise synchronization but requires waiting for the full response.

  • No real-time partial audio.
  • Full file returned upon completion.
  • Suitable for batch processing, not live broadcasting.

No SSML support

The engine does not interpret SSML tags for pauses, breaks, or emphasis. Voice control is limited to the 8 preset voices and the speed parameter (0.25–4.0).

If you need precise timing, adjust the speed value or split your text into smaller sentences to control pacing manually.

  • No SSML tags (e.g., <break>, <emphasis>).
  • Control speed via the speed field (0.25–4.0).
  • Use voice presets for tone variations.

Getting started

Sound wave illustration for: Getting started

To get started with DubSync API, sign up for an account on /get-api-key/ using Google or email. You will receive a $0.50 free trial credit valid for 7 days, allowing you to test the service immediately without upfront payment.

Once registered, you can issue a unique API key from your dashboard. This key authenticates your requests to the cloud engine, enabling you to generate voiceovers for your localization projects quickly and efficiently.

Sign up for an account

Visit /get-api-key/ to create your account. You can sign in using Google for instant access or use an email and password. Each site maintains independent accounts, so ensure you are on playtk71.com.

Get your API key

After signing in, your dashboard displays your active API key and current balance. This key is required for all API calls. You can issue a new key at any time, which immediately invalidates the old one.

Make your first request

Use the provided curl example or OpenAI SDK to send a POST request to the speech endpoint. Include your API key in the Authorization header, specify the text input, and select a voice from the 8 presets.

Common questions

Sound wave illustration for: Common questions

You can start generating voiceovers immediately without entering a credit card. New accounts created via Google sign-in receive a $0.50 free trial, which covers roughly 50,000 characters valid for seven days. If you sign up with an email and password, you begin with zero credit. There are no monthly subscriptions or minimum spend requirements, so you only pay for what you use.

Generation speed is approximately real-time, processing about 25 characters of text per second. Short phrases may take a few seconds to synthesize. Under heavy load, requests are queued, so longer texts may take slightly longer. Failed requests incur no charge, ensuring you never pay for errors.

Is it free to start?

Yes. Accounts registered with Google sign-in receive a $0.50 trial credit (about 50,000 characters) valid for 7 days. Email-only accounts start with no free credit. You can top up with USDT on the TRON network whenever needed, starting from just $5.

The trial is ideal for testing voice quality and integration speed before committing to larger dubbing projects. There are no hidden fees or subscription locks.

  • $0.50 trial for Google sign-ins
  • 7-day validity
  • No credit card required

How fast is the generation?

Synthesis runs at roughly real-time speed (25 characters per second). Short clips complete in seconds, while longer texts process sequentially. During high traffic, requests may queue briefly.

This speed is sufficient for most localization workflows where you batch-process scripts rather than requiring instant real-time playback in a live studio.

Can I use it commercially?

Yes. The generated audio files are yours to use in commercial dubbing, videos, or products. The input text is not used for training, and generated audio is not stored after the response.

Since voices are based on licensed recordings (LibriTTS-R) and public domain readings, you can use the output commercially without additional royalty tracking for standard usage.

Note that each file contains an inaudible watermark marking it as synthetic speech, which is standard for TTS but worth noting for your quality control process.

Support and contact

Sound wave illustration for: Support and contact

We keep things simple: all support is handled via email. There is no phone number or live chat to keep you waiting in a queue. Our team focuses on resolving your technical questions efficiently without the overhead of real-time support channels.

For quick answers, refer to our comprehensive documentation. We provide detailed guides on integration, pricing, and voice selection to help you get started faster.

Email support only

Reach out directly via email for any account, billing, or technical issues. We do not offer a phone number or live chat. This ensures your query is documented and addressed by a specialist rather than a generalist agent.

Documentation available

Our docs cover everything from basic curl examples to advanced integration tips. We believe clear documentation reduces the need for support tickets. Use the /docs/ page to find code snippets and endpoint details before emailing us.

No live chat

Unlike many SaaS platforms, we do not offer live chat. This allows us to maintain deep focus on each ticket. You will receive a thoughtful, detailed response rather than a quick, generic chat reply.

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. Locators master cultural nuance while dubbing artists breathe soul into characters for global audiences.

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. Python documentation
  4. curl documentation
  5. ISO 639 language codes
  6. W3C WAI: making audio and video accessible
  7. openai-python on GitHub
  8. LibriTTS-R paper (arXiv 2305.18802)
  9. MDN: Web Speech API
  10. LibriVox public-domain audiobooks
  11. LibriTTS-R corpus (OpenSLR 141)
  12. IETF RFC 7845: Ogg Opus

More on this site