Frequently asked questions
Who owns the VerbaLingua?
VerbaLingua is an independent storefront operating at play99ab.com. We are not affiliated with OpenAI, ElevenLabs, or any other major tech brand. Our team focuses on providing a straightforward text-to-speech API for product teams, allowing you to compare costs and switch providers without breaking your pipeline.
Are the voices really public domain?
Our voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and public-domain audiobook readings from LibriVox. This foundation ensures high-quality, legally clear synthetic speech for your applications without the complexity of managing individual voice licenses or dealing with restrictive usage terms.
Do you sell user data?
No. Input text is used only to produce the audio and is not used for training. Generated audio is returned in the response and not stored. We keep request metadata like time, characters, and status solely for billing purposes, ensuring your data remains private and secure.
Can I use the audio for podcasts?
Yes, you can use the generated audio for podcasts, apps, or any other purpose. Each file carries an inaudible watermark marking it as synthetic speech. Each file carries an inaudible watermark marking it as synthetic speech.
Is there a refund policy?
Failed requests are free, and the charge is returned in full if synthesis fails. There are no monthly plans or subscriptions to cancel. Your balance does not expire, and you only pay for successful character generation. You can issue a new API key anytime if needed.
How do I contact support?
Support is available by email only. We do not offer a phone number or live chat. For product teams needing practical assistance, email is our primary channel. We focus on delivering a stable API experience rather than managing multiple support touchpoints for every customer.
Are you affiliated with OpenAI?
No, we are not affiliated with OpenAI. Our API endpoint mimics the OpenAI speech endpoint shape, so you can use the official OpenAI SDKs by simply changing the base_url. This compatibility helps you switch providers easily, but we are an independent entity with our own pricing and infrastructure.
Do you offer enterprise contracts?
No, we do not offer enterprise contracts, SLAs, or uptime guarantees. We operate on a pay-as-you-go model with no minimum spend or seats. This approach allows product teams to scale usage up or down without negotiating complex terms or committing to long-term service level agreements.
Is the API stable?
We do not provide an official SLA or uptime guarantee. Requests are processed sequentially, so under load, a request may wait in a queue. Typical speed is roughly real time for long texts. We prioritize consistent pricing and compatibility over guaranteed latency metrics for every request.
Do you have a physical address?
We do not provide phone support. As an independent digital storefront, we operate entirely online. Our support is handled via email, and our infrastructure is cloud-based. This lean structure allows us to keep costs low and focus on API reliability.
Who we are
VerbaLingua is an independent storefront built for product teams shipping multilingual applications. We operate as a dedicated interface for a single text-to-speech API, allowing you to compare costs and switch providers without breaking your pipeline. Our editorial team focuses on clear, practical documentation that helps engineering and product teams integrate audio generation efficiently into language learning and educational apps.
Independent storefront
We are a distinct entity managing our own branding and billing, separate from the underlying API infrastructure. This structure gives you direct clarity on pricing and usage without intermediary markup. You interact with our specific domain, but the engine remains consistent and reliable for your multilingual needs.
Editorial team focus
Our content is crafted by writers who understand the workflow of language learning products. We prioritize clear integration steps and honest limitations over marketing hype. This ensures your team spends less time deciphering documentation and more time building features that matter to your users.
No corporate backing
Unlike larger platforms, we do not have a parent company or venture capital backing driving acquisition strategies. This means our pricing and feature set are driven by sustainable operational costs rather than growth metrics. You get a stable, focused service without the risk of sudden ecosystem changes or acquisitions.
Voice licensing
VerbaLingua’s eight voice presets—Aria, Nova, Luna, Sage, Felix, Milo, Atlas, and Orion—are derived from licensed recordings. We do not design voices from scratch or use celebrity impersonations. Instead, we use licensed recordings from LibriTTS-R and LibriVox to generate speech.
LibriTTS-R source
Our voices are built on the LibriTTS-R dataset, which provides a robust foundation for natural-sounding speech synthesis. This dataset serves as the core material for our eight preset voices, ensuring high audio quality across all supported languages and styles.
- Based on the LibriTTS-R dataset
- Provides the foundation for all 8 voice presets
- Ensures consistent audio quality across languages
CC BY 4.0 license
The underlying dataset is licensed under Creative Commons Attribution 4.0 International. This standard license allows for the commercial use of the data. You can integrate these voices into your products with confidence.
- CC BY 4.0 International license
- Commercial use permitted
- No individual attribution required for end-users
Public domain readings
The recordings originate from LibriVox, a platform dedicated to public-domain audiobooks. These are readings of works whose copyright has expired, ensuring that the source material is free from additional licensing constraints. This approach allows us to offer reliable voices without the risk of third-party copyright claims.
- Recordings from LibriVox
- Public-domain audiobook readings
- Free from additional licensing constraints
Data policy
VerbaLingua processes your text solely to generate the requested audio file. We do not use your input text or the resulting audio for model training or to improve our voice engines. Your data is transient by design.
No training on input
The text you send via the API is used exclusively for synthesis. Once the audio is returned, the input text is not stored for training purposes. This ensures your content remains private and is not ingested into our voice models.
No audio storage
Generated audio files are returned directly in the response and are not retained on our servers. We do not store the audio files you generate, ensuring that your voice assets remain yours immediately after the request completes.
Metadata retention
We retain request metadata—including timestamp, character count, and status—to support accurate billing and usage tracking. This data is kept only as long as necessary to maintain your account history and resolve any billing discrepancies.
Technical limits
VerbaLingua delivers audio via standard HTTP responses, meaning you receive the complete file only after synthesis finishes. This approach fits batch processing and offline playback. For language learning apps, this means audio buffers must be handled client-side rather than streamed progressively.
The API follows a fixed request shape compatible with OpenAI SDKs but lacks custom extensions. You send text and receive a file; there are no intermediate chunks or partial audio streams during generation. This simplicity reduces integration complexity for product teams managing multilingual content pipelines.
No streaming
Audio is returned as a single file (mp3, wav, opus, flac, or pcm) after processing completes. Requests are processed one after another, so under load a request may wait in a queue. Typical speed is roughly real-time (25 characters per second), but short phrases take a few seconds.
No SSML
The API does not support Speech Synthesis Markup Language (SSML). You cannot inject pauses, adjust pronunciation, or modify emphasis via XML tags. Control is limited to the speed parameter (0.25–4.0) and voice selection. For precise timing, split your text into smaller segments and concatenate the resulting audio files.
- No SSML tags for pauses or prosody
- No phoneme dictionaries or custom lexicons
- Speed control only via numeric multiplier
No voice cloning
We do not offer custom voice cloning or the ability to upload your own voice samples. All audio is generated from eight licensed presets based on LibriTTS-R recordings. Each file carries an inaudible watermark marking it as synthetic speech. This ensures consistent quality and clear licensing without the complexity of training models per user.
- No uploading of custom voice samples
- No per-user voice training
- Fixed set of 8 licensed presets

Support and contact
VerbaLingua keeps communication simple and asynchronous. We believe your engineering team should get clear, documented answers without the noise of real-time interruptions. Our support is handled entirely via email, ensuring every query is tracked and searchable.
We do not maintain a phone line or live chat widget. This approach allows our small editorial and technical team to focus on building a reliable API rather than managing concurrent call queues. You will receive a thorough response from a human, not a bot.
Email support only
All support requests are directed to our support email address. Because we are an independent storefront, our support team has direct access to the API logs to help troubleshoot specific request failures.
No phone number
We do not offer a telephone support line. This keeps our overhead low, which helps us maintain competitive pricing for character usage.
No live chat
Unlike many enterprise providers, we do not use live chat widgets on our site. This ensures you can work without pop-ups or notification distractions.
Network independence
VerbaLingua operates as an independent storefront for the same underlying text-to-speech API used across our network. Each site functions as a separate commercial entity, meaning your account, key, and billing are specific to play99ab.com and do not transfer to other stores in the network.
While the core synthesis engine and API endpoints remain identical, each storefront maintains its own distinct branding and customer relationships. This structure allows teams to compare pricing and onboarding flows across different interfaces without relying on a single monolithic provider.
Independent storefront
Our account system is isolated to this domain. Signing up with Google or email on play99ab.com creates a unique identity that does not exist on other network sites. This ensures clear billing boundaries and prevents cross-contamination of usage data between different storefronts.
Same API as others
The technical interface is consistent across the network. You use the same POST /v1/audio/speech endpoint and request schema regardless of which site you signed up on. Switching between network sites requires only updating your API key, not your codebase.
- Identical JSON request shapes
- Same rate limits and pricing models
- Consistent voice and language availability
Unique branding
While the engine is shared, our visual identity and support channels are distinct. This separation allows us to tailor the user experience for language learning product teams without the overhead of a larger corporate structure, while keeping the technical reliability uniform.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Users need intuitive, adaptive lessons that fit seamlessly into their daily routines for effective language acquisition.
23 languages
Any voice speaks any of these languages.
