Cloud Text-to-Speech API for Synthetic Data Teams

By SynthCode editorial team · Updated 2026-11-18

This guide explains how SynthCode supports product teams generating large-scale synthetic speech datasets for machine learning. We provide a reliable, character-priced HTTP API that integrates seamlessly into your data ingestion and model training workflows. By using standard JSON requests via the OpenAI SDK or plain HTTP, you can generate consistent audio output for your models without relying on proprietary wrappers, ensuring broad compatibility across your engineering stack.

Readers will learn about our voice licensing from LibriTTS-R, our strict data privacy policy, and our technical infrastructure. This is an independent storefront for developers who need transparent pricing, sequential processing, and email support without the overhead of subscriptions or complex enterprise contracts. Our API accepts text and returns audio files, making it easy to automate speech synthesis at scale for ML training.

Cloud Text-to-Speech API for Synthetic Data Teams

Frequently asked questions

Who runs SynthCode?

SynthCode is an independent storefront for a cloud text-to-speech API. Each site in the network operates independently, providing access to the same underlying engine. We are not affiliated with OpenAI, ElevenLabs, Google, Amazon, Microsoft, or any other brand mentioned in comparisons. Our focus is delivering reliable synthetic speech datasets for machine learning workflows.

Are the voices in SynthCode licensed?

Yes. Our eight voice presets are based on licensed recordings from LibriTTS-R (CC BY 4.0), sourced from LibriVox public-domain audiobook readings. Every generated audio file carries an inaudible watermark that marks it as synthetic speech, ensuring clear attribution and compliance with the source licenses while maintaining high-quality narration for your datasets.

Is my input text used for training models?

No. Input text is used solely to produce the audio file and is not used for training models. We respect your data privacy by ensuring your source text remains private to your generation process. Only request metadata, such as time, character count, and status, is retained for billing purposes.

How is data stored after audio generation?

Generated audio is returned in the response and is not stored by our servers. We retain request metadata, including timestamp, character count, and status, strictly for billing and usage tracking. Your input text is processed to create the audio and is not archived for long-term storage or model training.

Does SynthCode have an SLA?

No. SynthCode does not offer an SLA, uptime guarantee, or enterprise contract. We provide a reliable API for product teams, but we do not guarantee specific availability metrics. Our service is designed for scalable synthetic data generation without the overhead of enterprise-level service level agreements or dedicated support tiers.

How can I contact support?

Support is available by email only. We do not offer a phone number, live chat, or on-premise installation options. For technical issues or billing questions, please reach out via the contact page. Our team handles all inquiries through email to ensure detailed and accurate responses to your specific concerns.

Is SynthCode affiliated with OpenAI?

No. SynthCode is an independent provider, clearly distinct from OpenAI and other major tech brands. Our API endpoint and request shape are compatible with the OpenAI speech endpoint, so you can use the official OpenAI SDKs by changing the base_url to https://api.playvn500.com/v1.

What licenses do the voices use?

The voices are based on licensed recordings from LibriTTS-R, which uses the CC BY 4.0 license, and public-domain audiobook readings from LibriVox. This combination ensures high-quality, legally sourced synthetic voices. Each generated file includes an inaudible watermark to identify it as synthetic speech, maintaining transparency and compliance with source licensing requirements.

Our Mission for ML Data Teams

Sound wave illustration for: Our Mission for ML Data Teams

SynthCode exists to serve product teams building synthetic speech datasets for machine learning. We provide a reliable, character-priced HTTP API that integrates seamlessly into your data ingestion and model training workflows. Our goal is to eliminate infrastructure friction so your team can focus on model performance rather than audio generation details.

By offering scalable access to high-quality synthetic voices, we enable teams to generate the volume of audio data needed for robust training without the complexity of managing voice infrastructure. Our platform is designed specifically for the demands of large-scale data pipelines.

Providing reliable synthetic speech data

Providing reliable synthetic speech data

We deliver consistent audio output for every request, ensuring that your synthetic datasets remain uniform in quality. Each generated file carries an inaudible watermark, marking it clearly as synthetic speech for your downstream processing needs. This reliability allows you to trust the integrity of your training data.

Our system processes requests sequentially, ensuring fair usage and predictable output. With rate limits set at 60 requests per minute per key, you can plan your data generation cycles with confidence, knowing that your throughput will remain stable under load.

  • Consistent audio quality across all generations
  • Inaudible watermarking for synthetic data identification
  • Predictable sequential processing for stable pipelines

Supporting product teams with scalable APIs

Supporting product teams with scalable APIs

Our API is designed to fit directly into your existing data pipelines. By using a standard HTTP/JSON interface, you can integrate us with any language or framework, not just Python. The endpoint structure mirrors the OpenAI speech endpoint, allowing you to use existing SDKs with a simple base URL change.

We support high-volume generation by allowing you to split texts into parts of up to 4096 characters and join the resulting audio files. This flexibility ensures that even long-form content can be processed efficiently, supporting the diverse needs of modern ML datasets.

  • Standard HTTP/JSON interface for any stack
  • OpenAI-compatible endpoint structure
  • Support for splitting and joining long texts

Focusing on transparency and simplicity

We believe in clear, straightforward pricing. At $10 per 1 million characters, you know exactly what you will pay. Failed requests are free, so you never pay for errors. There are no monthly plans, no seats, and no hidden fees. You can top up with USDT on the TRON network for automatic, instant credit.

Our documentation and API responses are designed to be easy to understand. We provide ready curl examples and clear error messages, reducing the time your developers spend debugging integration issues. We support you via email, ensuring that technical questions are handled by a responsive team.

  • Transparent character-based pricing
  • No hidden fees or subscriptions
  • Responsive email support for technical issues

Voice Licensing and Sources

Sound wave illustration for: Voice Licensing and Sources

SynthCode provides a transparent foundation for synthetic speech datasets. Our eight voice presets are built directly from licensed recordings rather than proprietary neural weights, ensuring clear lineage for data teams auditing their training corpora. The voices are based on LibriTTS-R (CC BY 4.0), offering a legally compliant starting point for large-scale generation without the ambiguity of black-box models.

Voices based on LibriTTS-R (CC BY 4.0)

Our engine uses audio sourced from the LibriTTS-R dataset, which is licensed under Creative Commons Attribution 4.0. This means the acoustic source material is openly licensed, allowing product teams to integrate synthetic speech into machine learning workflows with clear attribution standards. Unlike proprietary voice models that obscure their origins, our source material is verifiable and openly documented.

  • LibriTTS-R dataset as the primary source
  • CC BY 4.0 licensing for commercial use
  • Verifiable acoustic lineage for audit trails

Recordings from LibriVox public-domain audiobooks

The underlying recordings originate from LibriVox, a collection of public-domain audiobook readings. These recordings form the basis of our eight presets: Aria, Nova, Luna, Sage, Felix, Milo, Atlas, and Orion. Each voice is a preset derived from these licensed recordings, providing a consistent and repeatable source for your synthetic data generation pipelines.

Inaudible watermarking on all generated audio

Every generated audio file carries an inaudible watermark that marks it as synthetic speech. This feature helps data teams distinguish between human-recorded and AI-generated samples within their training sets. It provides a reliable method for labeling synthetic data, which is critical for evaluating model performance on mixed-source corpora.

Voice Licensing and Sources

Data Policy and Privacy

Sound wave illustration for: Data Policy and Privacy

SynthCode processes your text exclusively to generate the requested audio file. The input text is not repurposed for other uses and is strictly limited to the synthesis task for which it was submitted.

Input text used only for synthesis

We do not retain your input text beyond what is necessary to generate the audio. The data flows through our systems solely to produce the audio response.

No training on user input data

Unlike many consumer platforms, we do not use your input text to train our models. Your data remains yours, ensuring your proprietary datasets stay private and do not influence our base engine.

Audio returned in response and not stored

Generated audio is returned directly in the API response and is not stored on our servers after delivery. Only request metadata—such as timestamp, character count, and status—is kept for billing purposes.

Technical Infrastructure

Sound wave illustration for: Technical Infrastructure

The API is a cloud-based service accessible via POST https://api.playvn500.com/v1/audio/speech. It accepts JSON bodies with text up to 4096 characters and returns audio files directly. The endpoint supports standard formats like mp3, wav, opus, flac, and pcm, allowing seamless integration into Python data pipelines without additional infrastructure.

Cloud-based API endpoint

Requests are sent to the cloud endpoint and processed sequentially. Under load, a request may wait in a queue before generation begins. Typical speed is roughly real-time, but short phrases take a few seconds. This sequential processing ensures consistent output quality for dataset generation without complex infrastructure management.

Requests processed sequentially

Keys are limited to 60 requests per minute, or 10 per minute before the first top-up. This rate limit ensures fair usage and prevents burst traffic from overwhelming the service during high-volume synthetic data creation. Long texts should be split into parts of up to 4096 characters and joined by the client.

Rate limits ensure fair usage

  • 60 requests per minute per key (10 before top-up)
  • Sequential processing with queueing under load
  • 4096 character limit per request
Technical Infrastructure

Support and Contact

Sound wave illustration for: Support and Contact

We handle all customer inquiries exclusively via email. This approach allows us to maintain clear, documented communication trails for every support ticket, which is particularly useful when troubleshooting API integration issues or reviewing billing details.

Email support for all inquiries

All communications are managed through email channels. This ensures that every request is logged and can be referenced later if needed, providing a transparent record of our interaction with your team.

No phone or live chat

We do not offer phone lines or live chat widgets. By avoiding real-time voice support, we reduce operational overhead and keep our pricing straightforward. You can reach us at any time without waiting in a queue.

Responsive team for technical issues

Our team focuses on resolving technical questions and billing inquiries efficiently. We prioritize clear, written explanations to help your developers integrate the TTS API smoothly into their data pipelines without unnecessary delays.

Since we are an independent storefront, our support is dedicated solely to SynthCode users. We aim to provide accurate answers regarding rate limits, voice selection, and API endpoints without corporate bureaucracy.

Support and Contact

Why Choose SynthCode

Sound wave illustration for: Why Choose SynthCode

SynthCode operates as an independent storefront, ensuring we are not affiliated with OpenAI, ElevenLabs, or any other major brand. Our API endpoint and request shapes mirror standard conventions, allowing you to use official OpenAI SDKs by simply changing the base_url to https://api.playvn500.com/v1. This approach gives you familiar tooling without vendor lock-in or implied branding.

Independent storefront, not affiliated with OpenAI

We are a distinct entity providing a reliable TTS API for synthetic data generation. Our infrastructure is built specifically for scale, not as a subsidiary of a larger tech giant. This independence means our roadmap is driven solely by the needs of data teams, not by the broader consumer market trends of other platforms.

Transparent pricing without subscriptions

Pay only for what you generate: $10 per 1 million characters, with failed requests refunded in full. There are no monthly plans, no seats, and no minimum spend. You top up with USDT (TRC-20) starting at $5, and your balance never expires. This model aligns costs directly with your dataset production volume, eliminating waste from unused credits or rigid tier structures.

Built for Python developers and data teams

Our API is designed to integrate seamlessly into your data ingestion pipelines. With a simple POST request to /v1/audio/speech, you can programmatically generate audio files in formats like MP3, WAV, FLAC, OPUS, and PCM. The consistent JSON body structure allows for easy scripting in Python, enabling you to automate the creation of large-scale synthetic speech datasets for machine learning training with minimal overhead.

  • Standard JSON request body compatible with OpenAI SDKs
  • Support for MP3, WAV, FLAC, OPUS, and PCM formats
  • Easy integration into existing Python data pipelines
  • Character-based pricing scales linearly with dataset size

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. High quality synthetic speech data enhances the accuracy and naturalness of our machine learning models.

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. ISO 639 language codes
  3. TRON developer hub: TRC-20
  4. OpenAI API reference: create speech
  5. Node.js documentation
  6. curl documentation
  7. Tether (USDT)
  8. W3C WAI: making audio and video accessible
  9. LibriVox public-domain audiobooks
  10. openai-python on GitHub
  11. IETF RFC 7845: Ogg Opus
  12. FFmpeg documentation

More on this site