Frequently asked questions
Is WebSpeech Notes affiliated with OpenAI?
No. We are an independent storefront for the same underlying text-to-speech API. While our endpoint structure mimics the OpenAI speech endpoint for compatibility, we are not affiliated with OpenAI, ElevenLabs, Google, Amazon, Microsoft, or any other brand mentioned in comparisons.
Who owns the voices used in the API?
The voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and public-domain audiobook readings from LibriVox. We use these licensed sources to provide eight distinct voice presets for your text-to-speech generation needs.
Is my data safe with you?
Input text is used only to produce the audio and is not used for training. We do not claim specific certifications like SOC 2 or HIPAA, but we keep request metadata (time, characters, status) strictly for billing purposes.
Do you store the audio files I generate?
No. Generated audio is returned in the response and not stored by us. We retain request metadata for billing, but the actual audio content you generate is not kept on our servers after delivery.
How do I contact support?
Support is available by email only. We do not offer a phone number, live chat, or SLA guarantees. You can reach us via the contact page or email link provided on the site.
Are the voices customizable?
No. We offer eight fixed presets (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion). There are no emotion tags, SSML support, or custom pronunciation lexicons. You cannot adjust pitch or tone beyond the speed setting.
Do you offer voice cloning?
No. We do not offer custom voice cloning, uploading your own voice, or voice design. The eight available voices are static presets derived from licensed recordings and cannot be modified or cloned by customers.
Is there a privacy policy?
Our data practices are defined by our facts: input text is used only for synthesis, not training, and audio is not stored. Metadata is kept for billing.
Can I request a new voice?
No. We do not offer voice design or custom voice creation. The eight presets are fixed. If you need a different voice, you must choose from the existing library of Aria, Nova, Luna, Sage, Felix, Milo, Atlas, or Orion.
How long do you keep my request metadata?
Request metadata, including time, characters, and status, is kept for billing purposes. We retain this data to verify usage and process charges accurately. Specific retention durations beyond billing needs are not specified.
Do you sell user data?
No. Input text is used only to produce the audio and is not used for training. We do not sell user data. We retain request metadata strictly for billing and usage verification.
Are the voices used for training?
No. The input text you send is used only to produce the audio and is not used for training. The voices themselves are based on existing licensed recordings from LibriVox and LibriTTS-R.
Is the API open source?
No. We do not provide an official SDK of our own. We recommend using the official OpenAI SDK or plain HTTP requests. We are a proprietary API service with specific rate limits and endpoints.
Where are the servers located?
We do not offer data-center choice or specify server locations. We do not provide on-premise installations or SLAs. The API is hosted in the cloud, but specific geographic locations are not disclosed.
Who We Are
WebSpeech Notes is an independent editorial team dedicated to helping indie developers build reliable voice features into their web applications. We focus on practical comparisons, clear pricing breakdowns, and straightforward integration guides that cut through the noise of modern API markets.
Our storefront offers a single, consistent text-to-speech API without the complexity of enterprise contracts or SLAs. We provide the tools developers need to move from prototype to production with transparency and precision.
Independent editorial team
We operate as a standalone entity, free from the marketing agendas of larger tech conglomerates. Our content is written by developers, for developers, ensuring that every guide and comparison reflects real-world usage scenarios rather than promotional talking points.
We maintain editorial independence, allowing us to present honest limitations alongside strengths. This approach helps you make informed decisions about your tech stack without hidden biases.
- Editorial independence
- Developer-written content
- Transparent pricing models
Not affiliated with major brands
We are not affiliated with OpenAI, ElevenLabs, Google, Amazon, or Microsoft. While we may reference their products for context or comparison, our API and infrastructure are entirely separate. We do not use their branding, infrastructure, or endorsement.
Our API endpoint, voices, and pricing structure are unique to WebSpeech Notes. This distinction ensures that your integration relies solely on our service, with clear lines of support and accountability.
Focused on indie developers
We understand the constraints of indie development: limited budgets, small teams, and a need for quick, reliable integrations. Our API is designed to be simple to implement, with clear documentation and predictable pricing.
We prioritize features that matter to small teams, such as straightforward character counting, easy top-ups via USDT, and a generous free trial. We avoid enterprise bloat, offering exactly what you need to get started and scale.
Our API
Our API follows a standard cloud pattern: send a POST request with your text, and receive an audio file. The endpoint structure mirrors the OpenAI speech API, allowing you to use official OpenAI SDKs simply by adjusting the base URL. This reduces integration friction for developers already familiar with common TTS workflows.
Cloud-based speech synthesis
The service processes requests sequentially, meaning audio generation happens on our servers rather than in the browser. This ensures consistent quality without draining the user’s device battery. You can control the output format, speed, and language directly via JSON parameters.
8 distinct voice presets
We offer eight voice models, each with a specific tonal character. Female voices include Aria (warm narrator), Nova (bright), Luna (soft storyteller), and Sage (calm instructor). Male options are Felix (energetic promo), Milo (friendly conversational), Atlas (neutral news), and Orion (deep documentary).
- 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
23 supported languages
The engine supports 23 languages, including English, Spanish, German, Japanese, and Arabic. Any voice can speak any supported language. If the language parameter is omitted, the system guesses the language from the script, defaulting to English for Latin text. Non-English Latin scripts should explicitly set the language code.
Voice Licensing
WebSpeech Notes provides an independent storefront for a cloud text-to-speech API, offering indie developers a transparent alternative to major platform offerings. Our editorial focus remains on practical integration guides and clear pricing breakdowns for building voice features into web applications.
We are not affiliated with OpenAI, ElevenLabs, or any other brand. Our team consists of developers who understand the specific needs of small-scale projects, emphasizing honest limitations and straightforward documentation over marketing hype.
LibriTTS-R source
Our eight voice presets—ranging from the warm narration of Aria to the deep documentary tone of Orion—are derived from licensed recordings. These voices are not artificially generated from scratch but are based on high-quality source data from the LibriTTS-R dataset.
CC BY 4.0 license
The source recordings are licensed under Creative Commons Attribution 4.0 International. This ensures that the foundational audio data used for our synthesis engines carries a clear, permissive license that allows for commercial use with proper attribution.
LibriVox public domain
Additional voice models utilize readings from LibriVox, which consists of public domain audiobooks. This dual-source approach ensures legal clarity and high-quality audio for our users, combining the structured dataset of LibriTTS-R with the natural variety of public domain narration.
Data Policy
WebSpeech Notes prioritizes your data privacy with a straightforward retention policy. We process your input solely to generate the requested audio file, ensuring that your source text is not stored for training or other purposes once the synthesis is complete.
The generated audio is returned directly in the response and is not persisted on our servers. This minimizes storage overhead and ensures your content remains transient unless you explicitly save it to your own infrastructure.
Input text not stored
Your input text is used exclusively for the immediate generation of the audio file. It is not retained for machine learning model training or repurposed for other services, ensuring your creative content remains distinct from our training datasets.
This approach maintains the integrity of your original inputs, treating them as ephemeral data necessary only for the single synthesis operation.
Audio not stored after generation
Once the audio file is delivered to your client via the API response, it is not stored on our servers. This ensures that your generated voice assets remain under your control without lingering copies in our infrastructure.
This policy reduces storage costs and simplifies data management for indie developers who prefer to host their own audio assets.
Metadata kept for billing
We retain request metadata, including timestamp, character count, and status, strictly for billing purposes. This data allows us to accurately calculate charges based on the $0.01 per 1,000 characters rate and verify successful transactions.
This metadata does not include the audio content or the original text, ensuring your core content remains private while we maintain accurate billing records.

Honest Limits
We provide a reliable service without the complexity of enterprise contracts. You get straightforward access to the API without hidden tier restrictions or mandatory support agreements.
No SLA guarantees
We do not offer SLAs, uptime guarantees, or enterprise contracts. Our service is designed for indie developers who need flexibility rather than rigid corporate commitments.
No uptime promise
There is no formal uptime promise. While we strive for consistency, unexpected outages can occur, and we do not compensate for downtime periods.
Queue processing under load
Requests are processed sequentially. During high traffic, your synthesis job may wait in a queue before starting. This ensures fair resource distribution but increases latency slightly.
Contact Us
We prioritize asynchronous communication to keep our editorial focus sharp. All support requests are handled via email, ensuring every question receives a thoughtful, documented response without the noise of real-time interruptions.
This approach aligns with our indie developer audience, who often prefer clear, written trails over fleeting conversations. You can reach us directly for technical queries or billing issues.
Email support only
Our primary channel is email. We review all incoming messages and respond with detailed solutions or updates. This ensures you have a permanent record of the support provided.
We do not promise immediate replies. However, we strive to resolve issues efficiently within standard business hours.
No live chat
Unlike many SaaS platforms, we do not offer live chat widgets. This keeps our overhead low and allows us to pass savings on to developers via our straightforward pricing model.
If you need urgent clarification, email remains the most reliable way to get a complete answer rather than a fragmented chat transcript.
No phone number
We have no dedicated phone line. Voice calls often lack the precision needed for technical API troubleshooting, where code snippets and error logs are essential.
Our email support allows you to attach logs, share curl examples, and receive structured responses that are easier to reference later than a verbal conversation.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Creating games alone is hard work but the freedom to follow your creative vision makes every late night worth it.
23 languages
Any voice speaks any of these languages.
Sources and further reading
- Wikipedia
- LibriTTS-R corpus (OpenSLR 141)
- Tether (USDT)
- IETF RFC 7845: Ogg Opus
- curl documentation
- Creative Commons Attribution 4.0
- openai-python on GitHub
- LibriVox public-domain audiobooks
- openai-node on GitHub
- W3C WAI: making audio and video accessible
- Python documentation
- LibriTTS-R paper (arXiv 2305.18802)
