Frequently asked questions
What is the free text to audio API rate limit?
The API limits requests to 60 per minute per key. Before your first top-up, the limit is 10 requests per minute. This applies to all accounts regardless of plan, ensuring consistent throughput for your language learning application's audio generation needs.
Which voices are best for language learning?
All eight voices support all 23 languages. Aria offers a warm, clear narrator style, while Sage provides a calm, instructional tone. Milo is friendly and conversational. You can test any voice with any language.
How many languages does the API support?
The API supports 23 languages: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. Any voice can speak any of these languages with equal quality.
Is there a free trial for the API?
Yes, signing in with Google provides $0.50 of credit, valid for 7 days. This covers approximately 50,000 characters. Accounts created with email and password do not receive this trial credit. You can start generating audio immediately to test quality and integration.
Can I use the API for commercial apps?
Yes, LinguaFlow provides a reliable text-to-speech API suitable for commercial language learning platforms and edtech products. There are no monthly plans, seats, or minimum spend requirements. You pay only for the characters you generate, with no complex infrastructure constraints.
What audio formats are available?
You can generate audio in mp3, wav, opus, flac, or pcm formats. Choose the format that best fits your application's playback requirements. All formats are generated from the same engine, ensuring consistent voice quality regardless of the output format selected.
How fast is the audio generation?
Generation is roughly real time for long texts, processing about 25 characters per second. Short phrases may take a few seconds. Requests are processed sequentially, so under load, a request may wait in a queue. This ensures stable performance for your app.
Do I need to specify the language?
If you do not set the language, the system guesses it from the script. Latin text defaults to English. For non-English Latin-script text, you should explicitly set the language parameter to ensure accurate pronunciation and intonation for your language learning content.
Are the voices based on real people?
Yes, the voices are based on licensed recordings from LibriTTS-R (CC BY 4.0), sourced from LibriVox public-domain audiobook readings. Each generated file contains an inaudible watermark marking it as synthetic speech.
Can I adjust the speaking speed?
Yes, you can adjust the speed between 0.25 and 4.0. This allows you to slow down audio for beginners or speed it up for advanced learners. The speed adjustment applies to all voices and languages, giving you precise control over the output.
What happens if a request fails?
If synthesis fails, the charge is returned in full. You are not billed for failed requests. This ensures you only pay for successfully generated audio files. You can retry the request or adjust parameters without financial penalty if the initial attempt fails.
Is there a minimum spend required?
No, there is no minimum spend required. You pay $0.01 per 1,000 characters. There are no monthly plans or seats. You can top up with as little as $5 using USDT on the TRON network, and your balance does not expire.
Do you offer voice cloning?
No, we do not offer custom voice cloning. You cannot upload your own voice or design a new voice. The API provides eight preset voices that can speak any of the 23 supported languages.
How is billing handled?
Billing is based on characters generated, at $10 per million characters. Payments are made via USDT on the TRON network (TRC-20). Top-ups are credited automatically within minutes. Failed requests are free, and your balance never expires. No invoices or VAT are involved.
First request
curl
curl https://api.play79y.com/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.mp3API for Language Learning Engines
LinguaFlow delivers precise synthetic speech for language learning platforms, supporting 23 languages including Arabic, Chinese, and English. Each voice is engineered for clarity, ensuring accurate pronunciation drills and reading comprehension exercises without complex infrastructure setup.
Natural pronunciation for 23 languages
Our API handles diverse linguistic scripts with high fidelity. You can generate audio for any of the 23 supported languages, such as Hindi, Japanese, or Turkish, using a single endpoint. The system automatically detects language from the input text if not specified, though explicit language codes are recommended for non-English Latin scripts.
- Supports 23 languages including Arabic, Chinese, Danish, Dutch, English, French, German, Greek, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish.
- Automatic language detection from script; explicit ISO codes recommended for non-English Latin text.
Fast generation for interactive exercises
Text is processed in near real-time, generating approximately 25 characters of audio per second for long texts. Short phrases may take a few seconds due to queue processing. This speed allows for responsive interactive exercises where users hear immediate audio feedback after typing.
- Typical speed: ~25 characters per second for long texts.
- Short phrases take a few seconds; requests are queued under load.
Consistent voice across lessons
Select from 8 distinct voice presets, such as Aria for warm narration or Atlas for neutral news tones. All voices are derived from licensed recordings, ensuring consistent quality. Each generated file includes an inaudible watermark to identify synthetic speech, maintaining brand integrity across your educational content.
- 8 voice presets including Aria (warm), Felix (energetic), and Atlas (neutral).
- Inaudible synthetic watermark in every audio file.
Supported Voices for Education
LinguaFlow provides eight distinct voice presets, each optimized for specific educational contexts. These voices are derived from licensed LibriTTS-R recordings and support 23 languages, ensuring consistent audio quality across diverse language learning exercises.
Aria: Warm narrator for reading
Aria offers a warm, clear narrative tone ideal for reading comprehension passages and story-based drills. Her consistent pacing supports learners in developing listening skills without distraction.
Felix: Energetic for promo content
Felix delivers an energetic, promotional tone suitable for marketing audio within edtech platforms or engaging introductory lessons. His bright delivery helps maintain user attention during shorter audio clips.
Atlas: Neutral tone for news
Atlas provides a neutral voice. Sage provides a calm, instructional tone. Atlas is suitable for news reading exercises or accessibility features where clarity is paramount.
Developer Integration
LinguaFlow integrates seamlessly into existing edtech stacks using standard HTTP clients. By adjusting the base_url to https://api.play79y.com/v1, the official OpenAI SDKs function without modification, allowing product teams to leverage familiar code patterns for audio generation.
OpenAI SDK compatibility
Use the official OpenAI SDK by pointing it to our API base. This eliminates the need for custom wrappers, enabling rapid integration of TTS features into language learning applications with minimal engineering overhead.
Simple JSON request format
Send a single POST request to /v1/audio/speech with a JSON body. The structure mirrors the OpenAI speech endpoint, requiring only model, input, and voice fields to generate audio instantly.
- Supports
tts-1andtts-1-hdmodels - Handles up to 4096 characters per request
- Optional
languageparameter for precise control
Multiple audio output formats
Specify the desired output format via the response_format field. Supported formats include MP3, WAV, OPUS, FLAC, and PCM, allowing flexible storage and playback strategies depending on your app’s network and device constraints.
Language Support Details
LinguaFlow supports 23 languages with high-fidelity synthesis, enabling precise pronunciation drills and immersive reading comprehension exercises. Each voice can render any supported language, ensuring consistent audio quality across your curriculum regardless of the target tongue.
Arabic, Chinese, Danish, Dutch
Support includes complex script handling for Arabic and tonal precision for Chinese, alongside clear articulation for Nordic and Western European languages. Set the language ISO code explicitly to avoid default English guesses for Latin-script inputs.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German
- Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian
English, French, German, Greek
Core languages like English, French, and German benefit from natural prosody using the tts-1 or tts-1-hd models. Greek adds another layer of linguistic diversity for your educational content without requiring separate voice models.
Hindi, Italian, Japanese, Korean
Support extends to Hindi, Italian, Japanese, and Korean, allowing you to build localized content quickly. All voices are interchangeable; use any preset voice to generate audio for any of these 23 supported languages.
Pricing Model
Usage-based billing charges $0.01 per 1,000 characters, with spaces and punctuation counted. Failed requests incur no cost, ensuring you only pay for successful synthesis.
Pay per character usage
Each character in your input text contributes to the total count. This linear pricing model scales efficiently with your application’s growth, eliminating unpredictable monthly fees.
- $0.01 per 1,000 characters
- Spaces and punctuation count
- Failed requests are free
Free trial credit included
Sign in with Google to receive $0.50 in trial credit, valid for 7 days. This allows you to generate approximately 50,000 characters before committing to a paid account. Email-based accounts start without trial credit.
- $0.50 trial credit for Google sign-ins
- Valid for 7 days
- Approximately 50,000 characters
Transparent USDT payments
Top-ups use USDT (TRC-20) on the TRON network. Minimum top-up is $5. Credits are added automatically within minutes. No invoices or VAT are applied.
- USDT (TRC-20) on TRON
- Minimum top-up $5
- Automatic credit within minutes
Technical Specifications
The API accepts input text up to 4096 characters per request. For longer content, split the text into segments and join the resulting audio files.
Adjust playback speed between 0.25x and 4.0x to suit different listening exercises. The engine supports MP3, WAV, OPUS, FLAC, and PCM output formats.
4096 character input limit
Each API call processes a maximum of 4096 characters. Spaces and punctuation count toward this limit. For texts exceeding this length, split the input into parts and concatenate the generated audio files.
- Maximum input: 4096 characters
- Characters include spaces and punctuation
- Split and join for longer texts
Speed adjustment from 0.25x to 4.0x
Control the generation speed using the speed parameter, accepting values from 0.25 to 4.0. This allows precise tuning for pronunciation drills or rapid listening exercises without altering the voice characteristics.
- Range: 0.25x to 4.0x
- Adjustable per request
- No voice quality degradation
MP3, WAV, OPUS, FLAC, PCM support
Select the response_format from MP3, WAV, OPUS, FLAC, or PCM. Choose formats based on your bandwidth requirements and playback compatibility. MP3 offers compression, while WAV and FLAC provide lossless quality.
- Formats: mp3, wav, opus, flac, pcm
- Lossless and compressed options
- Configurable per request

Voice Licensing
Our voices are generated from the LibriTTS-R dataset, which consists of public-domain audiobook recordings sourced from LibriVox. This foundation provides a natural, human-like quality while maintaining clear, consistent articulation ideal for language learning exercises.
The underlying recordings are licensed under CC BY 4.0. This transparent licensing model ensures that your educational content remains compliant without hidden royalty fees, allowing you to scale your pronunciation drills and reading comprehension tools with confidence.
LibriTTS-R dataset basis
The voices are based on licensed recordings from the LibriTTS-R dataset, sourced from LibriVox public-domain audiobook readings. This dataset ensures consistent acoustic properties across all eight voice presets, reducing the cognitive load for learners by providing predictable pronunciation patterns.
- Derived from public-domain LibriVox readings
- Optimized for clear, standard pronunciation
- Consistent acoustic properties across voices
CC BY 4.0 licensed recordings
All source recordings are licensed under Creative Commons Attribution 4.0 International (CC BY 4.0). This allows you to generate audio content for your platforms. You should still comply with the CC BY 4.0 license requirements for the source recordings.
- Standard CC BY 4.0 license
- No per-file attribution required
- Commercial use permitted
Inaudible synthetic watermark
Every audio file generated by the API includes an inaudible watermark that marks the content as synthetic speech. This feature helps platforms distinguish between human-recorded and AI-generated audio, aiding in content auditing and copyright management.
Performance Metrics
Synthesis operates at approximately real-time speed for extended passages, generating roughly 25 characters of text per second of audio. Short phrases may require a few additional seconds for processing overhead.
Real-time speed on long texts
Longer texts are split into 4096-character segments and joined seamlessly. This ensures consistent playback duration matching the spoken content without significant buffering delays.
Queued processing under load
Requests are processed sequentially. Under high traffic, a request may wait in a queue before synthesis begins. This ensures stability but may increase latency during peak usage periods.
Consistent latency standards
Latency remains predictable based on text length and current queue depth. Failed requests incur no charge and return instantly, avoiding wasted compute time for your application.
The API does not support streaming, so the full audio file is returned in a single response.
- Processing is sequential, not parallel.
- Failed requests are free and instant.
- No streaming or partial audio support.
Use Cases in EdTech
LinguaFlow enables scalable audio generation for educational software, allowing product teams to integrate synthetic speech directly into their learning engines. By supporting 23 languages with consistent voice presets, developers can automate content creation without managing complex infrastructure or licensing deals.
Automated pronunciation feedback
Generate precise audio examples for specific phonemes or words to help learners compare their speech against native-like pronunciation. The API’s consistent voice quality ensures that students hear uniform articulation across thousands of drill instances.
- Consistent articulation across all generated samples
- Ideal for phoneme-level repetition drills
- Supports 23 languages for global curricula
Digital textbook audio
Convert reading comprehension passages into audio files instantly, enabling students to listen to texts while reading. With a 4096-character limit per request, long texts are split and joined seamlessly, maintaining narrative flow without manual editing of audio segments.
- Text length up to 4096 characters per request
- Seamless joining of segmented audio files
- Support for multiple output formats like MP3 and WAV
Accessibility for visual impairment
Provide high-quality synthetic voiceovers for educational materials, ensuring compliance with accessibility standards. The API’s neutral and clear voices, such as Atlas for news or Sage for instructional content, deliver clear auditory access to course materials without the variability of human recording sessions.
- Clear, neutral voices for clarity
- Scalable generation for large content libraries
- No need for studio recording schedules
Billing and Limits
LinguaFlow charges $0.01 per 1,000 characters, with a free trial credit of $0.50 for Google sign-ups. Balances never expire, and failed requests incur no charge.
Payments are processed via USDT (TRC-20) with a $5 minimum. We do not offer subscriptions or invoices, keeping the model simple for your engineering team.
60 requests per minute limit
Each API key supports 60 requests per minute, scaling to 10 per minute before the first top-up. Longer texts are split into 4,096-character chunks and joined automatically.
- 60 req/min per key
- 10 req/min before first top-up
- Automatic chunking for long texts
Failed requests are free
If synthesis fails, the character count is deducted from your balance and returned in full. You only pay for successfully generated audio files.
No minimum spend required
There are no monthly plans, seats, or minimum spend requirements. Top up only when needed, and start generating audio immediately with your trial credit.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Mastering language learning requires consistent engagement and adaptive strategies for product teams.
23 languages
Any voice speaks any of these languages.
Sources and further reading
- Wikipedia
- IETF RFC 7845: Ogg Opus
- W3C WAI: making audio and video accessible
- FFmpeg documentation
- openai-python on GitHub
- LibriTTS-R corpus (OpenSLR 141)
- Tether (USDT)
- curl documentation
- LibriTTS-R paper (arXiv 2305.18802)
- LibriVox public-domain audiobooks
- Creative Commons Attribution 4.0
- OpenAI API reference: create speech
