Curisum API: Licensing, Data Policy & Support

By Curisum editorial team · Updated 2026-11-18

This page explains how Curisum Audio works for indie developers building digital news platforms. We handle the speech synthesis so you can focus on delivering articles in 23 supported languages, including English, Spanish, French, German, Japanese, and Hindi. You will find clear details on voice licensing, data usage, and billing metadata.

We are an independent storefront, not affiliated with OpenAI, ElevenLabs, Google, Amazon, Microsoft, or any other brand mentioned in comparisons. Support is by email only, with no live chat or phone lines. Read on to understand our data policy, technical limits, and how to get help with your TTS API integration.

Curisum API: Licensing, Data Policy & Support

Frequently asked questions

Who runs Curisum Audio?

Curisum Audio is an independent storefront for a text-to-speech API, built specifically for indie developers creating digital news platforms. We handle the speech synthesis infrastructure so you can focus on delivering news to your audience in 23 supported languages, without needing complex backend setups.

Where do the voices come from?

Our eight voice presets are based on licensed recordings from LibriTTS-R, which originates from LibriVox public-domain audiobook readings. These recordings form the foundation of our synthetic voices, maintaining clear licensing origins for the underlying data used in our models.

Is LibriVox public domain?

Yes. Our voices are based on the LibriTTS-R dataset, sourced from LibriVox public-domain audiobook readings. The dataset is licensed under CC BY 4.0, providing a solid open foundation for our engine. Additionally, every generated audio file carries an inaudible watermark to clearly mark it as synthetic speech.

Do you use my text for training?

No. Input text is used only to produce the audio for your specific request. We do not use your submitted text for training our models. This ensures your content remains private and is not repurposed to improve our underlying speech synthesis algorithms or datasets.

Do you store my audio files?

Generated audio is returned in the response and not stored by us. Once you receive the audio file, we do not keep a copy of the generated content. Only request metadata, such as time, characters, and status, is retained for billing and usage tracking purposes.

How long do you keep logs?

We keep request metadata, including time, characters, and status, for billing purposes. This data allows us to accurately track your usage and charges. We do not store the actual audio content or the input text beyond the immediate generation process, keeping your data minimal.

How can I contact support?

Support is available by email only. We do not offer phone support or live chat. You can reach our team via the contact page or support email. We aim to provide helpful, direct assistance to indie developers using our API for their news platforms.

Is there a phone number?

No, there is no phone number for Curisum Audio. All support interactions are conducted via email. This approach allows us to maintain a focused, asynchronous support structure that fits well with the workflows of indie developers and small teams using our text-to-speech API.

Do you have a privacy policy?

We do not currently host a separate privacy policy document. Our data practices are straightforward: input text is used solely to generate audio and is not used for training. Generated audio is returned in the response and not stored. We retain only request metadata (time, characters, status) for billing purposes.

Is this an official OpenAI product?

No, Curisum Audio is not affiliated with OpenAI. Our API uses the same request shape as the OpenAI speech endpoint, so their SDKs work by changing the base URL. However, we are an independent storefront providing our own text-to-speech API with our own voices, pricing, and infrastructure.

Who owns the generated audio?

Generated audio files include an inaudible watermark to identify them as synthetic speech. You may use the generated audio freely. We do not store the generated audio, and input text is used only for synthesis, not for training. Request metadata is kept for billing purposes. For specific usage rights and restrictions, please review our full terms and conditions.

Can I use voices commercially?

Yes. Our voices are based on LibriTTS-R (CC BY 4.0) recordings from LibriVox public-domain audiobook readings. You can integrate the generated audio into your digital news platforms. Note that every generated file carries an inaudible watermark marking it as synthetic speech.

Are the voices real people?

No, the voices are synthetic. They are based on licensed recordings from LibriTTS-R (CC BY 4.0), derived from LibriVox public-domain audiobook readings. While rooted in human speech, they are processed through our text-to-speech engine rather than being direct recordings. Additionally, every generated audio file carries an inaudible watermark that clearly marks it as synthetic speech.

Do you have a team size?

We are an independent storefront focused on providing a reliable text-to-speech API. We do not disclose specific team size metrics. Our structure is optimized to support indie developers creating digital news platforms, ensuring efficient infrastructure management and direct support through our email-based customer service channel.

Who We Are

Sound wave illustration for: Who We Are

Curisum Audio is an independent storefront for a single text-to-speech API. We are not affiliated with OpenAI, ElevenLabs, or any other brand. Our goal is to provide indie developers with a reliable API that turns articles into natural audio without requiring complex infrastructure.

Editorial team background

Editorial team background

We built this API to handle the technical needs of publishing. Our focus is on delivering a straightforward, consistent experience for news platforms, allowing you to generate audio from text without managing voice synthesis servers.

Focus on indie developers

Focus on indie developers

We designed the API for indie developers creating digital news platforms. By handling the speech synthesis, we let you focus on delivering content to your audience in any language, using standard HTTP requests and OpenAI-compatible SDKs.

Independent storefront

Curisum Audio operates as an independent storefront. We provide direct access to the TTS API with transparent pricing and simple account management, without the complexity of enterprise contracts or multi-brand affiliations.

Voice Licensing

Sound wave illustration for: Voice Licensing

Curisum Audio provides a reliable text-to-speech API for indie developers building digital news platforms. We are an independent storefront, not affiliated with major cloud providers. Our editorial team focuses on helping developers turn articles into natural audio without complex infrastructure.

LibriTTS-R source

Every voice in our library is based on licensed recordings from the LibriTTS-R dataset. These recordings originate from LibriVox public-domain audiobook readings. This foundation ensures that our synthetic voices carry the natural cadence and texture of human speech.

CC BY 4.0 license

The source data is available under the CC BY 4.0 license. This permits us to use these recordings to train our synthesis models while respecting the original creators' attribution rights. You receive the generated audio file without needing to manage these underlying license terms for your end users.

LibriVox public domain

LibriVox volunteers have dedicated years to reading public-domain books into the public domain. By leveraging this extensive archive, we access a diverse range of speaking styles and accents. This allows us to offer eight distinct voice presets that sound authentic and varied.

  • Volunteer-recorded audiobooks
  • Public domain source material
  • Diverse speaking styles
Voice Licensing

Data Policy

Sound wave illustration for: Data Policy

We process your input text strictly to generate the audio response. The text is not used for training our models, ensuring your article content remains private during the conversion process.

Input text usage

Your text is sent to the engine solely to produce the audio file. We do not repurpose your articles for model improvement or other purposes.

No training data

Unlike many AI services, we do not use your input text to train our TTS engine. Your articles are yours alone.

No audio storage

Generated audio is returned in the response and is not stored on our servers. Once you receive the file, our role ends.

Billing Data

Sound wave illustration for: Billing Data

We retain request metadata to ensure accurate billing and to process refunds for failed synthesis attempts. This data supports our transparent charging model without storing the generated audio itself.

Metadata retention

We keep logs of request time, character count, and status for billing purposes. This information is used solely to verify charges and process full refunds when synthesis fails.

Character counts

Your bill is based on the exact character count of your input text, including spaces and punctuation. This count is recorded for every request to calculate the $0.01 per 1,000 characters cost.

Request status logging

Each request’s status (success or failure) is logged. If a request fails, the associated charge is returned in full, ensuring you only pay for successfully generated audio.

Billing Data

Support Contact

Sound wave illustration for: Support Contact

Curisum Audio is an independent storefront for the same text-to-speech API used by other sites in our network. We are an editorial team focused on indie developers building digital news platforms. This page outlines how we handle your data, license our voices, and provide support.

Email only support

All communication happens via email. There are no phone numbers, live chat widgets, or ticketing portals to navigate. Simply send your message to our support address, and a member of our team will handle it directly.

No live chat

We do not offer live chat support. If you need immediate assistance during your development workflow, you can check our documentation or review the API status. For all other inquiries, email is our single, dedicated channel.

Response expectations

We review emails regularly and respond as quickly as possible. While we do not guarantee a specific response window, you can expect a reply within a reasonable timeframe on standard working days.

  • Email is the only support channel.
  • No phone or live chat available.
  • Direct team communication.
Support Contact

Technical Limits

Sound wave illustration for: Technical Limits

As an independent storefront, we operate without an Enterprise-grade Service Level Agreement or a formal uptime guarantee. Our infrastructure is optimized for indie developers, meaning we prioritize stable synthesis over redundant, costly redundancy. While we strive for consistent availability, unexpected outages may occur without prior notice or compensation.

No SLA guarantee

We do not offer a formal SLA with financial credits for downtime. This keeps our pricing simple and aligned with indie budgets, but it means you should build retry logic into your application rather than relying on a guaranteed 99.9% availability contract.

No uptime promise

Our servers may experience brief interruptions due to maintenance or unexpected load. Since we are not bound by a strict uptime promise, critical production systems should implement their own fallback mechanisms or caching strategies to ensure smooth delivery to your audience.

Queue behavior

Requests are processed sequentially. Under heavy load, your synthesis job may wait in a queue before processing begins. Longer texts are split into parts of up to 4096 characters, which are then joined. This sequential processing ensures high quality but means latency can increase during peak usage periods.

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. news outlets often overlook the creative spirit of indie developers in their coverage.

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. Tether (USDT)
  3. LibriVox public-domain audiobooks
  4. Creative Commons Attribution 4.0
  5. ISO 639 language codes
  6. openai-python on GitHub
  7. LibriTTS-R corpus (OpenSLR 141)
  8. openai-node on GitHub
  9. LibriTTS-R paper (arXiv 2305.18802)
  10. Node.js documentation
  11. OpenAI API reference: create speech
  12. W3C WAI: making audio and video accessible

More on this site