Frequently asked questions
Who runs Lucky TTS Review?
Lucky TTS Review is an independent storefront for a synthetic speech API, serving startups building ML datasets. We focus on cost-per-character, language coverage, and data rights for training pipelines, providing expert analysis without affiliation to the underlying API providers.
Are the voices licensed for commercial use?
Yes. The voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and public-domain audiobook readings from LibriVox. This licensing structure supports commercial use in synthetic speech datasets for ML applications without additional voice-specific restrictions.
Is my input text used for training?
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, ensuring your proprietary data remains isolated from model training pipelines.
How long is my data stored?
Request metadata (time, characters, status) is kept for billing purposes. Input text is not stored after generation, and generated audio is returned in the response and not retained on our servers for long-term storage.
Can I contact you via phone?
No. Support is by email only. We do not offer a phone number or live chat. You can reach our editorial and support team via email for questions regarding API reviews, pricing, or technical details.
Do you offer enterprise support?
No. We do not offer enterprise contracts, SLAs, or uptime guarantees. Our service is designed for startups and independent developers using the API directly, with standard email support available for all account-related inquiries.
Are the reviews sponsored?
No. Lucky TTS Review is an independent storefront for a single API. We evaluate based on facts like cost-per-character and language coverage. We are not affiliated with OpenAI, ElevenLabs, or other brands mentioned in our comparisons.
How do you verify API facts?
We verify facts by directly testing the API endpoints and reviewing official documentation. Our analysis focuses on concrete metrics like rate limits, character counts, and voice presets, ensuring all reported figures match the actual API behavior.
Can I suggest a voice for review?
Yes. You can suggest voices or APIs via our contact page. We evaluate based on relevance to ML datasets and speech products, prioritizing factors like language coverage, cost efficiency, and commercial licensing for synthetic speech use cases.
Is the editorial team independent?
Yes. Lucky TTS Review operates as an independent storefront for a single API. Our content reflects the facts of this API. We maintain objectivity by focusing on measurable data points like pricing, speed, and technical specifications.
Do you accept advertising?
No. We do not accept advertising or sponsorships that influence our reviews. Our evaluations are based on factual analysis of the API’s performance, pricing, and data rights, ensuring unbiased recommendations for startups building speech products.
How often are reviews updated?
Reviews are updated when significant changes occur in the API’s pricing, voice list, or technical limits. We monitor rate limits, character counts, and supported languages to ensure our data remains accurate for your integration planning.
Can I use the reviewed APIs commercially?
Yes. The API allows commercial use, with voices based on CC BY 4.0 and public-domain sources. You can generate audio for ML datasets, with no restrictions on commercial application, provided you adhere to the API’s terms and pricing.
What is the data retention period?
Request metadata (time, characters, status) is kept for billing. Input text is not stored after generation, and generated audio is returned in the response and not stored. This ensures minimal data retention for privacy and compliance.
Who We Are
Lucky TTS Review provides independent analysis of synthetic speech APIs, specifically targeting startups building ML datasets and speech products. We evaluate technical performance, cost-per-character, and data rights to help engineering teams select the right infrastructure for their training pipelines.
Expert Review Team
Our analysis is conducted by engineers and data scientists with deep experience in audio processing and dataset curation. We understand the specific constraints of training models, from character limits to format requirements, ensuring our reviews reflect real-world development needs rather than marketing claims.
Focus on ML Startups
We prioritize metrics that matter for ML workflows: character-based pricing, language coverage, and the absence of training clauses in the data policy. Our reviews help startups avoid hidden costs and ensure their synthetic audio assets can be freely used in commercial models.
Independent Analysis
We are an independent storefront evaluating the same API from a reviewer’s perspective. Our findings are based on direct API usage, rate limit testing, and character counting, providing transparent figures without affiliation to the underlying provider or other major brands.
Voice Licensing and Sources
Our review of Lucky TTS highlights that the eight voice presets are derived from licensed recordings found in the LibriTTS-R dataset. This foundation allows startups to build synthetic speech products with known provenance, ensuring that the audio stems from verifiable sources rather than opaque neural network training data.
The voices include specific presets like Aria, Nova, and Felix, all sharing the same underlying data origin. Understanding this source is critical for teams managing ML datasets, as it directly impacts the licensing terms under which you can use the generated audio files in your own products.
LibriTTS-R Dataset
The voices are based on recordings from the LibriTTS-R dataset, which contains hours of read English speech. This provides a consistent baseline for the synthetic voices, ensuring uniform quality across the eight available presets without mixing disparate source materials.
- Based on standard read English speech
- Ensures consistent voice quality
- Transparent data origin for ML pipelines
CC BY 4.0 License Details
The underlying dataset is licensed under Creative Commons Attribution 4.0 International. This license applies to the dataset; generated audio files carry an inaudible watermark marking them as synthetic speech. For startups integrating these voices into commercial ML datasets, this license offers clear legal grounds for usage.
- Permits commercial use
- Requires attribution
- Allows derivative works
Public Domain Recordings
The recordings originate from LibriVox public-domain audiobook readings. Since the source material is in the public domain, the copyright restrictions are minimal, primarily governed by the CC BY 4.0 license applied to the dataset. This simplifies rights management for generated audio outputs.
- Source material is public domain
- Simplified rights management
- Ideal for scalable dataset creation
Data Policy and Privacy
We evaluate synthetic speech APIs for startups building ML datasets and speech products. Our reviews focus on cost-per-character, language coverage, and data rights for training pipelines.
Understanding how your input data is handled is critical for ML workflows. We ensure transparency regarding text usage, storage policies, and metadata retention to help you build reliable, compliant speech products.
Input Text Usage
Input text is used exclusively to produce the audio file for that specific request. It is not used for training our models or improving the TTS engine, ensuring your proprietary data remains private during synthesis.
Audio Storage Policy
Generated audio is returned directly in the API response and is not stored on our servers after delivery. This ensures that your final audio assets remain under your control without lingering copies in our infrastructure.
Billing Metadata Retention
We retain request metadata—including timestamp, character count, and status—strictly for billing and usage verification purposes. This data supports accurate invoicing and helps you audit your API consumption without storing the actual audio content.
Technical Limitations and Honesty
We evaluate synthetic speech APIs for startups building ML datasets and speech products. Our reviews focus on cost-per-character, language coverage, and data rights for training pipelines.
No Streaming or SSML
The API returns complete audio files, not partial streams. Requests are processed one after another, so under load a request may wait in a queue. There is no WebSocket support for real-time audio delivery, and SSML tags are not parsed.
- Fixed response formats: mp3, wav, opus, flac, pcm
- No phoneme dictionaries or custom pronunciation lexicons
No Custom Cloning
We use eight fixed presets based on licensed recordings from the LibriTTS-R dataset and LibriVox public-domain audiobook readings. You cannot upload your own voice or design custom tones. Every generated file carries an inaudible watermark marking it as synthetic speech.
No SLA Guarantees
There is no uptime guarantee or enterprise contract. Requests are processed sequentially, so under load, synthesis may wait in a queue. We do not offer on-premise installation or data-center choices, and support is limited to email.

Contact and Support
We handle all customer inquiries exclusively via email. There is no live chat widget, no phone number, and no ticketing portal for direct voice interaction. If you need assistance with your API key, billing discrepancies, or technical questions about the TTS endpoint, email is the single channel to reach our team.
Email Support Only
All support requests are managed through a dedicated email address. This ensures that every query is logged and can be referenced against your account usage and billing metadata if needed. Please include your API key in the subject line to help us locate your account quickly.
No Live Chat or Phone
We do not offer real-time voice support or instant messaging widgets on this site. This decision aligns with our focus on providing a straightforward, infrastructure-first experience for ML startups who prefer asynchronous, documented communication over immediate verbal resolution. You will not find a phone number or a chat bubble on any page.
Response Expectations
Support is by email only. We do not guarantee a specific SLA. While we aim to resolve billing and key-related issues promptly, response times may vary depending on the volume of requests. For urgent API issues, check the status of your requests via the dashboard before emailing.
Editorial Standards
We evaluate synthetic speech APIs for startups building ML datasets. Our reviews focus on cost-per-character, language coverage, and data rights for training pipelines.
We measure concrete metrics like characters per second and exact pricing tiers, ensuring every figure matches current API documentation.
Fact-Based Reviews
Every claim is tied to verifiable API behavior, such as the 4096-character limit or the 60 requests-per-minute rate cap. We cite exact endpoints and response formats rather than general impressions.
No Hype or Superlatives
We avoid terms like "best" or "fastest" unless defined by specific benchmarks. Descriptions focus on observable traits, such as the warm, clear narrator style of the 'aria' voice or the deep, documentary tone of 'orion'.
Transparent Comparisons
When comparing voices or languages, we list exact support levels, such as the 23 supported languages or the absence of Vietnamese. All comparisons include the date of verification to ensure accuracy.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Additionally, their licensing terms for synthetic data are less transparent than some alternatives.
23 languages
Any voice speaks any of these languages.
Sources and further reading
- Wikipedia
- LibriVox public-domain audiobooks
- W3C WAI: making audio and video accessible
- Tether (USDT)
- Node.js documentation
- MDN: Web Speech API
- OpenAI API reference: create speech
- LibriTTS-R corpus (OpenSLR 141)
- Creative Commons Attribution 4.0
- Python documentation
- IETF RFC 7845: Ogg Opus
- LibriTTS-R paper (arXiv 2305.18802)
