Frequently asked questions
Is AI-generated voice considered voice cloning?
No. Our service uses eight licensed presets derived from LibriTTS-R and LibriVox recordings. It does not perform custom voice cloning or allow you to upload your own voice. Each generated file carries an inaudible watermark marking it as synthetic speech, distinct from cloned voices.
Do I need consent to use synthetic voices in my courses?
The voices are based on licensed recordings (LibriTTS-R CC BY 4.0) and public-domain audiobook readings. While input text is not used for training, you should verify your specific usage rights. We do not offer custom cloning, so consent for specific individuals is not required.
How is voice cloning different from preset voices?
Voice cloning creates a unique model from your voice data. We use eight fixed presets (e.g., Aria, Felix) based on existing licensed recordings. We do not offer custom voice cloning or voice design. You select from these eight options for consistent, compliant audio.
Can I clone my own voice for LMS content?
No. We do not offer custom voice cloning or allow you to upload your own voice. You must choose from our eight preset voices (e.g., Luna, Atlas). This ensures consistent quality without the legal complexities of cloning specific individuals' voices.
How much does it cost to generate audio for large courses?
Pricing is $0.01 per 1,000 characters. Failed requests are free. You can start with a $0.50 trial credit. Top-ups via USDT (TRC-20) start at $5. There are no monthly plans, seats, or minimum spending requirements. Balance does not expire.
Are there limits on the number of characters I can generate?
Each request accepts up to 4,096 characters. For longer texts, split them into parts and join the audio files. You can send 60 requests per minute. Short phrases take a few seconds; long texts generate at roughly real-time speed.
Which languages are supported for international courses?
We support 23 languages, including English, Spanish, German, French, Japanese, and Hindi. Any voice can speak any supported language. If not set, the language is guessed from the script. Vietnamese, Thai, and Ukrainian are not supported.
Is the audio output watermarked?
Yes. Every generated file carries an inaudible watermark that marks it as synthetic speech. This helps identify the content as AI-generated without affecting the audio quality or the listener's experience during course playback.
Can I use the generated audio in commercial LMS platforms?
Yes. The voices are based on licensed recordings (LibriTTS-R CC BY 4.0) and public-domain readings. Input text is used only to produce audio and is not used for training. You can integrate the audio directly into your LMS.
Do I need technical knowledge to integrate the API?
The API follows the OpenAI speech endpoint shape. You can use official OpenAI SDKs by changing the base URL. The /get-api-key/ page provides a ready curl example.
How fast is the text-to-speech generation?
Generation is roughly real-time for long texts (25 characters per second). Short phrases take a few seconds. Requests are processed sequentially, so under load, a request may wait in a queue. Typical speed is efficient for bulk course production.
What happens if a request fails? Am I charged?
Failed requests are free. If synthesis fails, the charge is returned in full. You only pay for successful audio generations. This ensures you are not billed for errors or issues during the text-to-speech process.
Can I change my API key if it is compromised?
Yes. You can issue a new API key at any time. The old key stops working immediately. This allows you to secure your account quickly if a key is compromised without losing your balance or usage history.
Is there a minimum spending requirement?
No. There are no monthly plans, no seats, and no minimum spend. You can top up with a minimum of $5 via USDT. Your balance does not expire, and you pay only for the characters you generate.
First request
curl
curl https://api.c168s.info/v1/audio/speech \
-H "Authorization: Bearer $TTS_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"tts-1","voice":"aria","input":"Hello from the API.","response_format":"mp3"}' \
--output hello.mp3Understanding Synthetic Voice Technology
The API converts text into audio files via a simple HTTP request. Unlike streaming services, it processes requests sequentially, returning a complete audio file rather than partial streams. This predictable workflow suits batch processing in learning management systems.
What is AI Voice Generation?
Synthetic speech uses neural networks to read text aloud. It does not clone specific human voices but relies on licensed recordings from LibriTTS-R and LibriVox. Each file contains an inaudible watermark identifying it as synthetic.
How Text-to-Speech Works in LMS
LMS teams send text to a POST endpoint and receive an audio file. Long scripts are split into 4096-character segments, joined into one file. This ensures consistent delivery without complex infrastructure.
The system handles up to 60 requests per minute. Typical speed mirrors real-time speech, taking about one second per 25 characters for short phrases.
Voice Cloning: Consent and Licensing
Synthetic speech generation relies on licensed recordings rather than direct voice cloning. Our engine uses audio from LibriVox public-domain audiobook readings, which are released under a CC BY 4.0 license. This means the base voice recordings are licensed under CC BY 4.0, simplifying licensing for educational use.
Unlike custom voice cloning services that require specific consent agreements for unique voice profiles, our preset voices are standardized assets. LMS teams can deploy these voices across thousands of lessons without worrying about individual licensing expirations or renegotiating rights for specific narrators.
Who Owns the Voice?
The underlying voice models are derived from public-domain recordings licensed under CC BY 4.0. When you generate audio, you receive a synthetic file marked with an inaudible watermark to identify it as AI-generated. You retain full rights to the generated audio files for your LMS content.
- Voices based on LibriTTS-R (CC BY 4.0) public-domain recordings
- Generated files carry an inaudible synthetic watermark
- No individual performer rights to manage per lesson
Consent in Educational Content
Since our voices are presets and not custom clones of specific living individuals, explicit consent from real people is not required for standard course production. This simplifies compliance for large educational libraries where hundreds of narrators might otherwise be involved. The synthetic nature of the audio is clear, avoiding impersonation claims while maintaining high production value.
Preset Voices for Course Narration
Narrata provides eight distinct voice presets designed for educational narration, ranging from warm narrators to calm instructional guides. These voices are based on licensed recordings from LibriTTS-R and LibriVox.
Using a consistent voice across all modules helps learners focus on content rather than adapting to different speaking styles. Each voice supports any of the 23 available languages, allowing L&D teams to maintain a uniform brand experience across global educational libraries.
8 Professional Preset Voices
The API offers eight presets, including female voices like Aria (warm narrator) and Luna (soft storyteller), and male voices like Atlas (neutral news) and Orion (deep documentary). Teams can select the tone that matches their course style, from upbeat promos to instructional guides.
Every generated file includes an inaudible watermark marking it as synthetic speech, providing a clear distinction from human-recorded audio while maintaining high quality.
- 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
Consistency Across Modules
Consistency is critical for large course libraries. By applying the same voice preset across all modules, teams ensure that learners do not experience jarring shifts in tone or pace as they progress through a curriculum.
This uniformity supports better retention and reduces cognitive load, as the auditory experience remains predictable. The API’s straightforward model selection makes it easy to deploy the same voice across thousands of lessons.
Why LMS Teams Choose API Integration
API integration connects the TTS engine directly to your learning management system, allowing dynamic audio generation for course modules without manual file uploads. This ensures consistent narration across all content updates and reduces the administrative burden of managing large audio libraries.
Direct LMS Integration
By using the standard OpenAI SDK with a modified base URL, teams can embed synthesis calls directly into their LMS workflows. This method supports the same request structure as the OpenAI speech endpoint, making it easy to integrate existing tools.
- Standard JSON request body compatible with OpenAI SDKs
- Simple base URL configuration for existing integrations
- Automatic handling of authentication via Bearer tokens
Automation for Bulk Content
LMS teams can automate the generation of thousands of lessons by splitting texts into 4096-character chunks and processing them in sequence. This approach handles large educational libraries efficiently, ensuring that every module receives high-quality narration without manual intervention or significant delays.
- Automatic text splitting for long scripts
- Sequential processing to maintain audio continuity
- Scalable volume for entire course catalogs
Cost Efficiency at Scale
Narrata’s pricing model charges $10 per million characters, equating to $0.01 per 1,000 characters. This direct cost structure allows L&D teams to forecast audio production expenses precisely based on course length and character count.
Pay Per Character Model
Every character, including spaces and punctuation, counts toward the total. Failed requests are refunded in full, ensuring you only pay for successfully generated audio. This efficiency supports large-scale content libraries without waste.
- $0.01 per 1,000 characters
- Spaces and punctuation count
- Failed requests are free
No Monthly Fees
There are no monthly plans, seats, or minimum spend requirements. Balance does not expire, and there are no subscriptions or invoices with VAT. You top up via USDT on the TRON network, with a minimum of $5 credited automatically within minutes.
- No monthly plans or seats
- Balance does not expire
- Top-up via TRC-20 USDT

Quality and Performance
Generation speed averages roughly real-time, producing approximately 25 characters of text per second for long texts. Short phrases may take a few seconds, and requests are processed sequentially, so heavy load can introduce queue delays.
Real-Time Synthesis Speed
Long-form content, such as multi-module courses, typically generates at a pace close to natural speaking time. However, because requests are handled one after another, high concurrent volume may result in temporary queuing. Users should anticipate slight delays during peak usage periods rather than instant delivery.
Audio Format Options
The API supports five standard audio formats: MP3, WAV, OPUS, FLAC, and PCM. Teams can select the format that best fits their LMS player capabilities, balancing file size against quality requirements. All generated files include an inaudible watermark to identify synthetic speech.
Supporting Multiple Languages
The API supports 23 languages, including English, Spanish, German, French, Japanese, and Chinese. Any voice can speak any of these languages, allowing you to maintain consistent character tones across diverse linguistic content without hiring multiple voice actors.
23 Languages Available
We support Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. For non-English Latin-script text, set the language parameter to ensure accurate pronunciation, as the system defaults to English otherwise.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek
- Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese
- Russian, Spanish, Swahili, Swedish, Turkish
Global Course Reach
Use preset voices that naturally handle multiple languages. This ensures your L&D teams can scale course production across regions quickly, with each generated file marked as synthetic speech for clear compliance.
Data Privacy and Usage
Input text is used solely to generate the target audio file and is not retained for training future models. This ensures that your specific course content remains distinct from the general data pool used to refine the underlying synthesis engine.
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 strictly for billing and usage verification purposes.
No Training on Your Data
Our processing pipeline treats your input text as a transient request. Once the audio file is generated and delivered, the original text is not archived for model improvement or retraining.
This separation ensures that your proprietary course scripts and educational materials do not influence future voice synthesis outcomes, maintaining the consistency of your brand's audio identity.
Secure Generation
Audio files are returned in the API response and are not stored on our servers. This minimizes the attack surface for your content.
We retain only the minimal metadata required for accurate billing: the request timestamp, character count, and status code. No audio content or source text remains on our servers after the transaction completes.
Getting Started with Synthetic Audio
Narrata eliminates friction for L&D teams by offering a $0.50 free trial credit, valid for seven days, when you sign in with Google. This allows your technical leads to validate integration quality without immediate financial commitment or credit card details. Accounts created via email do not receive this trial credit.
The API follows the OpenAI standard, meaning your existing engineering resources can begin integration immediately using familiar tools. The base URL is https://api.c168s.info/v1, allowing you to leverage standard HTTP clients or the official OpenAI SDK by simply adjusting the base endpoint configuration.
Free Trial Credit
Start with a $0.50 trial credit (approx. 50,000 characters) valid for 7 days upon Google sign-in. This provides a risk-free environment to test voice quality and API responsiveness before committing to a top-up via USDT (TRC-20). No monthly fees or subscriptions are required to begin testing.
- Valid for 7 days
- Requires Google sign-in
- Approx. 50,000 characters
Developer Documentation
Integration is straightforward using standard HTTP POST requests. The endpoint https://api.c168s.info/v1/audio/speech accepts JSON bodies identical to the OpenAI format. You can use the official OpenAI SDKs by changing the base_url to https://api.c168s.info/v1 and providing your unique API key.
Documentation details the required fields: model, input, voice, and response_format. You can also retrieve available voices and formats via the GET /v1/voices endpoint. Authentication uses a Bearer token, and you can rotate keys instantly if needed.
- OpenAI SDK compatible
- Standard JSON body
- Rotate keys instantly
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. This preset is particularly effective for keeping learners engaged in short modules or promotional training content.
23 languages
Any voice speaks any of these languages.
