Frequently asked questions
What is a text-to-speech API?
It is a cloud service where you send text via a POST request and receive an audio file. You use the endpoint https://api.cb125.info/v1/audio/speech with your API key. It works like the OpenAI speech endpoint, allowing you to use official OpenAI SDKs by changing the base URL.
Why use TTS for teaching Japanese?
FableSays provides a reliable Japanese text reader API with clear voices, supporting lesson planning and accessibility. You can generate audio from texts up to 4096 characters, helping students hear correct pronunciation. It is practical for creating audio materials for children’s stories and fairy tales without manual recording.
Which voice is best for stories?
Q: Which voice is suitable for stories? A: Luna is a soft, female voice ideal for fairy tales. Sage offers a calm, instructional tone for educational content. Both are available presets that can speak Japanese. Choose Luna for narrative engagement or Sage for clear explanations in your Japanese language lessons. Any voice can speak any of the supported languages, so you may also experiment with other presets to find the best fit for your specific lesson style.
How do I specify Japanese language?
Include the language field in your JSON body with the ISO code for Japanese. Although the API guesses language from script, Latin text defaults to English. Explicitly setting the language ensures correct pronunciation for Japanese text, preventing the system from defaulting to English for any Latin-script elements.
Can I change the speed of the audio?
Yes, you can adjust the speed parameter between 0.25 and 4.0. This allows you to slow down speech for beginners or speed it up for advanced students. You can fine-tune the audio generation to match your students' listening comprehension levels.
Does the API support SSML?
No, the API does not support SSML. You cannot use phoneme dictionaries or custom pronunciation lexicons. You must rely on the standard text input and voice presets. If you need specific pronunciation, you may need to adjust the input text or choose a different voice preset.
How much does it cost per lesson?
The cost is $10 per 1 million characters, or $0.01 per 1,000 characters. Spaces and punctuation count. For a typical lesson of 1,000 characters, it costs $0.01. Failed requests are free, so you only pay for successful audio generation. There are no monthly plans or subscriptions.
Can I use the API on mobile devices?
Yes, you can use the API on any device with internet access. Since it is a standard HTTP API, you can integrate it into mobile apps or web views. There is no dedicated mobile SDK, but you can use the official OpenAI SDK or plain HTTP requests.
Is the audio copyrighted?
The voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and LibriVox public-domain audiobooks. Each generated file carries an inaudible watermark marking it as synthetic speech. The input text is not used for training.
How do I handle long texts?
Split your text into parts of up to 4096 characters per request. Send each part to the API and join the resulting audio files. This ensures each request stays within the character limit. You can process these sequentially or in parallel, though requests are processed one after another under load.
What if the pronunciation is wrong?
The API uses preset voices based on licensed recordings. There is no custom voice cloning or phoneme dictionary to correct specific pronunciations. If pronunciation is incorrect, you may need to adjust the input text or try a different voice preset. Support is available via email if issues persist.
Can I use this for accessibility?
Yes, FableSays supports accessibility by providing clear voices for Japanese text. You can generate audio for students with reading difficulties or visual impairments. The API is simple to integrate, allowing you to add audio versions of lesson materials easily. It helps make educational content more inclusive.
How do I integrate with my LMS?
Use the standard HTTP endpoint or OpenAI SDK to send text and receive audio. You can create a simple script to generate audio files and upload them to your LMS. Since there is no dedicated plugin, you will need to handle the API calls and file management manually.
What is the rate limit?
The limit is 60 requests per minute per key. Before the first top-up, it is 10 requests per minute. If you exceed the limit, requests may wait in a queue. Ensure your integration handles rate limits gracefully to avoid delays in generating audio for your students.
Understanding TTS for Education
Text-to-speech technology transforms static lesson materials into accessible audio resources, allowing students to listen while reading. This dual-modality approach reinforces comprehension and retention, making it easier for learners to follow complex texts without straining their eyes.
Benefits of Audio in Learning
Audio files enable passive learning during commutes or breaks. Teachers can assign listening tasks that complement written exercises, ensuring students engage with the material in multiple ways. The API returns high-quality audio files quickly, supporting dynamic lesson planning.
- Reinforces reading through auditory cues
- Supports multitasking and flexible study habits
- Reduces visual fatigue for extended reading sessions
Accessibility for Diverse Students
Clear audio output supports students with dyslexia or visual impairments by providing an alternative to text-only formats. It also aids English language learners who benefit from hearing correct pronunciation before attempting to speak themselves.
- Supports visual impairment needs
- Aids dyslexic readers with auditory reinforcement
- Assists non-native speakers in pronunciation
Consistency in Pronunciation
Unlike human readers who may vary in tone or pace, the API delivers consistent pronunciation every time. This reliability ensures that students hear the same standard pronunciation repeatedly, which is crucial for language acquisition and memorization.
- Uniform voice quality across all lessons
- No variation in accent or tone
- Reliable for repetitive listening exercises
Choosing the Right Voice
Selecting the correct voice preset ensures your Japanese lessons resonate with students. Each voice offers a distinct tone, from the soft warmth of Luna to the calm clarity of Sage, allowing teachers to match audio to the lesson’s intent without guessing.
Luna for Storytelling
Luna delivers a soft, storytelling quality ideal for reading fairy tales or narrative texts. Her warm tone keeps students engaged during longer passages, making her a strong choice for immersive reading sessions.
Sage for Instructions
Sage provides a calm, instructional voice perfect for clear directions or grammar explanations. Her neutral yet friendly tone helps students focus on linguistic details without distraction, ensuring instructions are understood precisely.
Aria for Narration
Aria offers a warm, clear narrator style suitable for general lesson overviews. Her consistent pacing and clarity make her reliable for standard text-to-speech needs, ensuring every word is heard distinctly.
Technical Setup
To begin using FableSays for your Japanese lessons, you first need an API key. Visit our /get-api-key/ page and sign in using your Google account or create a new account with an email and password. Once logged in, your unique key will be displayed alongside your current balance and usage statistics.
Obtaining an API Key
Accounts are created per site, so ensure you are on the correct domain. After signing in, you will see your active key ready for use. You can issue a new key at any time, which immediately invalidates the previous one for security.
- Sign up via /get-api-key/ using Google or email.
- One active key per account; replace anytime.
- Key format starts with 'tts-...'
Configuring the Base URL
Configure your client to point to our API base URL: https://api.cb125.info/v1. This base URL handles all text-to-speech requests. You can use the official OpenAI SDKs by simply changing the base URL, or use standard HTTP libraries for more control.
- Base URL: https://api.cb125.info/v1
- Compatible with OpenAI SDKs (change base_url).
- Standard HTTP POST requests work as well.
Testing with CURL
Verify your setup with a simple command. Replace YOUR_API_KEY with your actual key and adjust the text as needed. This confirms connectivity and character counting before you generate full lesson audio.
- curl -X POST https://api.cb125.info/v1/audio/speech \ -H "Authorization: Bearer YOUR_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"tts-1","input":"こんにちは","voice":"luna","response_format":"mp3"}'
Creating Lesson Audio
Generating audio for Japanese lessons starts with clean, prepared text. Ensure your source material is free of complex formatting or markdown that the API cannot parse. Since the API processes text up to 4096 characters per request, preparing concise segments ensures reliable synthesis and faster generation times for your students.
You send these texts via a simple POST request to the speech endpoint. The API returns the audio file directly, which you can then embed into your lesson materials or learning management system. This straightforward workflow allows teachers to quickly produce consistent pronunciation audio without manual recording efforts.
Preparing Text Files
Before sending text, clean your Japanese scripts to ensure smooth synthesis. Remove special characters or complex line breaks that might confuse the engine. Keeping text under 4096 characters per segment helps maintain high-quality audio output and prevents request failures due to size limits.
Ensure your text files are saved in a standard encoding like UTF-8. This prevents character encoding issues that could alter pronunciation or cause silent audio files. Simple plain text formats work best for direct API integration.
- Use plain text or UTF-8 encoded files.
- Remove complex markdown or special symbols.
- Keep segments under 4096 characters.
Sending Requests
Use the API endpoint to send your prepared text. Include your API key in the authorization header and specify the desired voice, such as Luna for storytelling or Sage for instructional content. The system handles the synthesis and returns the audio file.
You can test the request using cURL or any HTTP client. The response will contain the audio data in your chosen format, ready for immediate use in educational materials.
- Include API key in header.
- Specify voice and language parameters.
- Choose output format (mp3, wav, opus).
Managing Character Limits
The API limits each request to 4096 characters. If your lesson text exceeds this, split it into smaller segments. Each segment generates a separate audio file, which you can later concatenate for a complete lesson.
Monitor your character usage to stay within budget. Each character counts, including spaces and punctuation. Understanding this limit helps you plan your audio generation efficiently, avoiding unnecessary charges or failed requests.
- Limit text to 4096 characters per request.
- Split long texts into segments.
- Count all characters, including spaces.
Handling Long Texts
Long texts like full fairy tales exceed the 4096-character limit per request. You must split the text into smaller parts and send each part individually to the API.
After receiving the audio files for each segment, join them in order. This ensures the complete story plays in order for students.
Splitting Text Automatically
Divide your lesson text into chunks of up to 4096 characters, counting spaces and punctuation. Send each chunk as a separate POST request to https://api.cb125.info/v1/audio/speech.
Keep track of the segment order. The API processes requests sequentially, so under load, earlier segments may finish before later ones, but you must maintain the original sequence.
- Count characters carefully; spaces and punctuation count toward the 4096 limit.
- Use a script to split text at sentence boundaries for natural pauses.
Joining Audio Files
Receive each audio file (mp3, wav, opus, flac, or pcm) in the response body. Save these files locally or on your server with sequential names.
Use an audio library or simple command-line tool to concatenate the files in the exact order they were generated. This creates a single continuous audio file for the lesson.
Ensure the format matches across all segments to avoid playback issues. MP3 and WAV are widely supported by learning management systems.
Ensuring Smooth Playback
Verify that the joined audio plays without clicks or silence between segments. If using PCM or WAV, bit-depth and sample rate should match.
Test the final file with a sample student to ensure clarity and continuity. Adjust speed if needed, but keep it consistent across all segments.
Remember that failed requests are free, so retrying a segment is safe if audio quality is poor.
- Test the final joined file before distributing to students.
- Retry failed segments immediately; they are credited back.
Adjusting Speed and Tone
To tailor audio for classroom use, adjust the speed parameter (0.25–4.0) in your request. This controls playback rate without altering voice quality, helping students process spoken Japanese at their own pace.
Speed Range 0.25–4.0
Set values between 0.25 and 4.0. Lower speeds slow narration for clarity; higher speeds accelerate for quick reviews. The API accepts any float in this range, allowing precise control over reading speed.
Slower Speed for Beginners
Use speeds around 0.75–0.9 for students new to Japanese. Slower pacing helps learners distinguish phonemes and follow story structures in children’s tales without feeling rushed.
Faster Speed for Review
Increase speed to 1.2–1.5 for students already familiar with the material. Faster playback reinforces listening skills and helps students practice real-time comprehension during lesson reviews.
Language Support
Accurate pronunciation is critical for language acquisition. Our API supports Japanese alongside 22 other languages, ensuring students hear clear, standardized speech for lessons and stories.
Japanese Script Support
The engine handles Japanese characters natively without requiring additional configuration. You can input text directly in Hiragana, Katakana, or Kanji, and the TTS will generate clear audio suitable for elementary and intermediate learners.
Latin Script Default
If you omit the language field, the system guesses based on the script. Latin text defaults to English, which may cause mispronunciation if you input Japanese using Romaji. Always specify the language for mixed-script texts.
Explicit Language Codes
For reliable results, always set the language parameter to ja when sending Japanese text. This ensures the voice models apply the correct phonetic rules, regardless of the input script format.

Cost Management
FableSays charges $0.01 per 1,000 characters, with spaces and punctuation counted. You can monitor your exact character usage and remaining balance directly on the dashboard after signing in with Google or email.
Tracking Character Usage
Access your real-time usage statistics via the account dashboard. The system tracks every request, allowing you to see exactly how many characters have been processed for each lesson.
Budgeting for Lessons
Since pricing is linear, you can easily estimate costs before generating audio. Keep texts under 4,096 characters per request to simplify billing, or split longer stories manually.
Using the Free Trial
Google sign-ups receive $0.50 in credit, covering approximately 50,000 characters for seven days. This allows you to test voice quality and integration without immediate payment.
- Credit expires after 7 days
- Email/password accounts start with zero balance
- No subscription required
Integration with LMS
Integrate FableSays with your learning management system by sending text payloads directly from your course materials. Use the OpenAI-compatible SDK to generate audio files that students can access via standard links, ensuring consistent pronunciation across all lesson modules.
Connecting to Learning Platforms
Configure your LMS to call the API endpoint for each lesson text. Pass the Japanese script and select a voice like Luna for stories or Sage for instructions. The system returns an audio file that you can embed directly into your course page.
- Use standard HTTP requests compatible with most LMS plugins
- Supports all major formats including mp3 and wav for broad player compatibility
Automating Audio Generation
Pre-generate audio for entire textbooks before the semester starts. Write a simple script to iterate through your lesson texts, send them to the API, and store the resulting audio files in your LMS media library. This ensures instant playback without server lag during class.
- Batch process up to 4096 characters per request
- Store generated URLs in your LMS database for quick access
API Webhooks
The API does not currently support webhooks for status updates. Monitor your usage dashboard or check request responses manually to confirm audio generation completion. For large batches, rely on the predictable real-time processing speed to estimate total generation time.
Troubleshooting
When generating audio for Japanese lessons, errors usually stem from invalid character counts or unsupported models. Ensure your input stays within the 4096-character limit per request to avoid truncation or silent failures. Verify that the model field is set to tts-1 or tts-1-hd and that your API key is correctly formatted in the authorization header.
Rate limits restrict you to 60 requests per minute. If you exceed this, requests will queue or fail. Failed requests are free, but successful ones consume your character balance. Monitor your usage dashboard to stay within limits, especially when generating audio for large class sets.
Common Request Errors
Check that the input text does not exceed 4096 characters. If it does, split the text manually or via your script before sending. Ensure the voice parameter matches one of the eight available presets like Luna or Sage. Invalid voice names result in immediate rejection.
- Ensure
inputlength ≤ 4096 characters - Verify
modelistts-1ortts-1-hd - Check
voicematches available presets (e.g., Luna, Sage)
Rate Limit Issues
You are limited to 60 requests per minute per key. New accounts start with 10 requests per minute. If you hit the limit, wait until the window resets or upgrade your balance. Requests processed but failed do not count against your limit if no audio is generated.
- Limit: 60 requests/minute (10 for new accounts)
- Failed requests are free
- Requests queue under load
Audio Format Checks
Select a supported format: mp3, wav, opus, flac, or pcm. If the audio plays incorrectly, verify the format matches your player’s capabilities. The response returns the audio file directly; ensure your code handles binary data correctly without corrupting the stream.
- Supported formats: mp3, wav, opus, flac, pcm
- Handle binary response data correctly
- Verify player compatibility
Best Practices
Following consistent preparation steps ensures reliable audio generation for classroom materials. Small adjustments to your text input and voice selection prevent common errors and save time during lesson planning.
Cleaning Text Before Input
Remove extra spaces, line breaks, and punctuation that might confuse the synthesizer. Ensure the text fits within the 4096-character limit per request to avoid silent truncation.
- Strip unnecessary whitespace
- Verify character count before sending
- Check for special symbols
Testing Voices First
Generate short samples with Luna for stories or Sage for instructions before committing to full lessons. This verifies pronunciation accuracy and tone suitability for your specific student group.
- Use 100-character test strings
- Compare Luna vs. Aria
- Check Japanese script rendering
Monitoring Usage
Track character consumption via the dashboard to stay within your $0.01 per 1,000 characters budget. Regular checks help avoid unexpected charges and ensure consistent supply for recurring classes.
- Check balance weekly
- Log usage patterns
- Top up before depletion
Support and Contact
Support for FableSays is handled exclusively via email. We do not offer a phone number, live chat, or SLA guarantees. For technical questions regarding the API or billing issues with your TRC-20 top-up, please email us directly. We respond to all inquiries as quickly as possible, though wait times may vary based on volume.
Email Support Only
All support requests go to our dedicated email address. Please include your API key and a description of the issue. Since we are a lean team focused on product reliability, email remains our primary channel for resolving account and integration queries.
Documentation Links
Most technical questions are answered in our comprehensive docs. The /docs/ page covers all endpoints, including the Japanese language support and voice presets. For quick code examples, refer to the /get-api-key/ page, which includes ready-to-use cURL commands for immediate testing.
- API Reference: https://cb125.info/docs/
- Quick Start: https://cb125.info/get-api-key/
- Pricing Details: https://cb125.info/pricing/
Community Resources
While we do not host an official forum, teachers often share lesson plans using our API. You can find community-driven tips on social media by searching for #FableSays or #JapaneseTTS. We encourage educators to share their workflows to help others integrate audio into their curricula more effectively.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. It works well for young adult fiction or modern stories where a traditional fairy tale voice might feel too juvenile.
23 languages
Any voice speaks any of these languages.
Sources and further reading
- Wikipedia
- TRON developer hub: TRC-20
- MDN: Web Speech API
- Node.js documentation
- W3C WAI: making audio and video accessible
- openai-python on GitHub
- OpenAI API reference: create speech
- Tether (USDT)
- Creative Commons Attribution 4.0
- ISO 639 language codes
- LibriTTS-R paper (arXiv 2305.18802)
- LibriTTS-R corpus (OpenSLR 141)
