Frequently asked questions
What is the character limit per request?
Each request accepts up to 4096 characters of input text. This limit applies to the entire text payload sent in a single API call. For longer manuscripts, you must split the content into chunks that fit within this constraint to ensure successful synthesis.
How do I split long texts?
Split your manuscript into parts of up to 4096 characters. Generate audio for each part individually using the API. Then, join the resulting audio files together to create a complete audiobook. This approach ensures each segment is processed successfully.
Do I need to code?
No coding is required. You can use standard HTTP tools or the official OpenAI SDKs by changing the base URL to https://api.playk74.com/v1. Alternatively, use the ready curl example on the /get-api-key/ page to start generating audio immediately without writing code.
Which voice is best for fiction?
Aria offers a warm, clear narrator style, while Luna provides a soft, storytelling tone. Nova is bright and upbeat, suitable for lighter fiction. Choose based on your book’s mood; all voices support 23 languages and adapt to your text’s script.
Can I change speed?
Yes, you can adjust the playback speed between 0.25 and 4.0. This allows you to slow down complex passages or speed up lighter content. The speed parameter is included in the JSON body of your POST request to the speech endpoint.
What is the best format?
There is no single best format; choose based on your workflow. MP3 offers broad compatibility, while WAV and FLAC provide higher fidelity. OPUS ensures efficient compression, and PCM is uncompressed. All formats are available via the response_format parameter in your API request.
How do I join audio files?
Generate audio for each text chunk separately. Then, use any standard audio editing software or command-line tools like ffmpeg to concatenate the resulting files. The API returns each file individually, so you handle the final assembly yourself.
Is the audio watermark audible?
No, the watermark is inaudible. Every generated file carries a digital watermark that marks it as synthetic speech. You can listen to the audio without hearing any artificial tones or clicks added specifically for identification purposes.
Can I edit the audio?
Yes, you can edit the generated audio files. Since the API returns standard audio formats like MP3, WAV, or FLAC, you can use any audio editor to trim, fade, or adjust the files before publishing your audiobook.
How long does generation take?
Generation takes roughly real-time for long texts, about 25 characters per second of audio. Short phrases may take a few seconds. Requests are processed sequentially, so under heavy load, a request might wait in a queue before starting.
What if a request fails?
If synthesis fails, the charge is returned in full. You are not billed for failed requests. Simply retry the request with the same text and parameters, or check the error response to identify any issues with the input or account.
Do I own the rights?
The API does not explicitly grant copyright ownership of the final audio product. You receive and control the generated files, which are not stored by us. Input text is used solely for synthesis and not for training. While voices derive from licensed and public-domain sources, the API provides the file without explicitly transferring IP rights to the resulting audio.
Can I use it for Audible?
Yes, you can use the generated audio for Audible. The API produces standard audio formats like MP3, WAV, and FLAC. Most services accept standard MP3 files, which Audible primarily supports. Since the output is synthetic speech with an inaudible watermark, you should verify Audible’s current disclosure requirements for AI-generated content before publishing.
How do I handle errors?
Check the HTTP status code and error message in the response. Common issues include invalid characters, exceeded limits, or authentication errors. Ensure your API key is correct and the text length is under 4096 characters. Failed requests are free.
What languages are supported?
We support 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.
Do spaces count as characters?
Yes, spaces and punctuation count as characters. The character limit of 4096 applies to the entire input string, including all spaces and special characters. Plan your text chunks accordingly to stay within the limit for each request.
Can I use my own voice?
No, custom voice cloning is not offered. You cannot upload or train your own voice. Instead, choose from eight preset voices: aria, nova, luna, sage, felix, milo, atlas, and orion. Each voice has a distinct tone, but the selection is limited to these specific options available in the API.
Is there a mobile app?
No, there is no mobile app. The service is a cloud API accessed via HTTP requests. You can use it on any device with an internet connection and an HTTP client, such as a browser or a script.
How do I pay?
Top up your balance using USDT on the TRON network (TRC-20). The minimum deposit is $5, and funds are credited automatically within minutes. We do not offer monthly plans, subscriptions, or credit cards. Your balance never expires, and you only pay for the characters you actually use.
Preparation
Preparing your manuscript for audiobook production starts with cleaning the text file. Remove extra spaces and line breaks to ensure smooth reading. Check the character count, as spaces and punctuation count toward the 4096-character limit per request. This ensures your text fits within a single API call, avoiding unnecessary splitting.
Clean your manuscript
Start by removing any non-standard characters or formatting that might confuse the TTS engine. Ensure the text is plain UTF-8 encoded text without markdown or HTML tags. This reduces the risk of pronunciation errors and ensures consistent output quality.
- Remove markdown symbols like ** or ##
- Convert special characters to standard letters
- Ensure plain UTF-8 encoding
Remove extra spaces
Extra spaces or line breaks can cause unnatural pauses in the generated audio. Use a text editor to consolidate multiple spaces into single spaces and remove blank lines. This creates a continuous flow of speech, mimicking natural reading rhythm.
- Replace multiple spaces with one
- Delete blank lines
- Ensure consistent spacing
Check character count
Each API request supports up to 4096 characters, including spaces and punctuation. Before sending your text, verify the character count to avoid truncation. If your manuscript exceeds this limit, you will need to split it into smaller segments for processing.
- Count all characters including spaces
- Verify against 4096-character limit
- Split text if necessary
Choose a Voice
Selecting the right voice is crucial for audiobook engagement. Our eight presets offer distinct tonal qualities, ranging from warm narrators to deep documentary styles. Each voice can speak all 23 supported languages, allowing you to match the vocal style to your manuscript’s genre without language constraints. Choose a preset that aligns with your story’s atmosphere to maintain listener interest throughout the project.
Match your voice to your content type. Use Aria or Luna for storytelling and romance, Felix for energetic promos, or Atlas for neutral news. Ensure the tone complements the narrative without overpowering it. Testing short samples helps you verify that the voice fits the mood before committing to a full manuscript.
Select a preset
Choose from eight distinct voices designed for specific narrative styles. Each voice is based on licensed recordings from LibriTTS-R (CC BY 4.0), derived from LibriVox public-domain audiobook readings. This ensures high-quality, consistent synthetic speech that sounds natural and engaging for long-form content.
All voices support all 23 languages. If your text uses Latin script but is not English, explicitly set the language parameter to avoid incorrect auto-detection. This ensures accurate pronunciation and pacing regardless of the voice selected.
- 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
Match tone to genre
Align your voice choice with your book’s genre to enhance listener immersion. Use Orion for deep, documentary-style non-fiction or Atlas for neutral, news-like narratives. For lighter content, consider Nova’s upbeat tone or Milo’s conversational style. Consistency in tone helps maintain a professional audio product.
Avoid switching voices mid-book unless it serves a specific narrative device. Stick to one preset for the majority of your text to ensure a cohesive listening experience. This consistency is key for no-code authors who want a polished result without complex editing.
Test short samples
Generate short samples using the API to test voice quality before processing your entire manuscript. Send small text chunks to verify pronunciation, pacing, and emotional tone. This step helps you identify any potential issues with specific words or phrases that might sound unnatural.
Check the generated audio for clarity and natural flow. Ensure the voice matches your expectations for the final product. If the tone feels off, try a different preset or adjust the speed parameter to find the best fit for your audience.

Split Long Texts
Split your manuscript into segments of up to 4096 characters per request. Each segment generates a separate audio file that must be joined later. This ensures reliable synthesis without hitting API constraints.
For texts exceeding this limit, divide the content into logical chunks. Avoid breaking sentences mid-word to maintain natural speech flow. Each part is processed independently, so plan your splits carefully before sending requests.
4096 character limit
The API accepts a maximum of 4096 characters per request, including spaces and punctuation. Texts longer than this must be divided into smaller parts. Each part generates a distinct audio file that you will combine later.
If you send a single request with more than 4096 characters, the system may reject it. To avoid errors, always split your manuscript into chunks that stay within this limit before submitting.
- Count every space and punctuation mark as a character
- Split at natural pauses like periods or paragraph breaks
- Keep chunks under 4096 characters to ensure reliable synthesis
Split strategy
Divide your manuscript into logical segments, such as chapters or scenes. This makes it easier to manage individual audio files and maintain consistent tone across your audiobook. Use a simple text editor to count characters before sending.
Avoid splitting mid-sentence to preserve natural speech flow. Each segment should be self-contained and clear. This approach ensures the TTS engine generates coherent audio without awkward pauses or cut-offs.
- Split at chapter or scene boundaries
- Count characters before sending each request
- Use a text editor to verify chunk length
Joining audio files
After generating audio for each segment, download the files and merge them using an audio editor. This creates a continuous audiobook file. Ensure consistent volume and tone across segments for a professional result.
You can use free tools like Audacity or online mergers to combine files. Maintain the same voice and speed settings throughout to keep the audiobook cohesive. This step ensures a seamless listening experience for your audience.
- Use Audacity or similar tools to merge files
- Keep voice and speed settings consistent
- Check audio quality before final export
Send Requests
To generate audio, send a POST request to https://api.playk74.com/v1/audio/speech with your API key in the Authorization header. The JSON body requires the model (tts-1 or tts-1-hd), input text, voice, and response_format. This endpoint mirrors the OpenAI speech API structure, allowing you to use existing SDKs by simply adjusting the base URL.
Use the API
Construct your request with the required fields: model, input (max 4096 characters), voice, and response_format. Include your Bearer token for authentication. The API processes requests sequentially, so under heavy load, your request may queue briefly before synthesis begins.
- Endpoint: POST /v1/audio/speech
- Auth: Bearer tts-... key
- Format: JSON body
Monitor progress
Since requests are processed one after another, you may experience delays during high traffic. The API returns the audio file directly upon successful completion. For long texts, you will receive multiple audio files corresponding to your split segments.
Handle failures
If synthesis fails for any reason, the request is free, and any charged amount is returned in full. Check your request payload for valid voice names and character limits. Retry failed requests after a short delay to avoid rate limiting.
- Failed requests are free
- Refund issued automatically
- Check character limits
Process Audio
Once your requests complete, download the generated audio files using the API response. Since the API processes requests sequentially, long manuscripts may require patience. Failed requests are free, so you only pay for successfully generated segments.
Download files
Retrieve each audio file immediately after the synthesis completes. The API returns the audio directly in the response, allowing you to save the file locally or to your preferred storage.
Merge segments
Combine the individual audio files into a single audiobook file. Since the API splits text into chunks of up to 4096 characters, you must join these segments to create a continuous listening experience.
- Use standard audio editing software to concatenate files.
- Ensure the order matches your manuscript structure.
Check quality
Listen to the merged file to ensure smooth transitions between segments. Verify that the voice tone remains consistent throughout the entire manuscript, as switching voices mid-chapter can disrupt the listener's immersion.
Export Formats
Selecting the right audio format depends on your workflow. The API supports MP3, WAV, Opus, FLAC, and PCM to match different distribution and editing needs.
MP3 for distribution
MP3 is the standard for audiobook platforms like Audible and Spotify. It offers a strong balance between file size and audio quality, making it ideal for efficient storage and fast streaming on end-user devices.
WAV for editing
Use WAV when you need uncompressed, high-fidelity audio for post-production. This format preserves all details from the TTS engine, allowing you to adjust pacing or apply effects without generational loss before final export.
Opus for web
Opus provides efficient compression with small file sizes, reducing bandwidth usage for web embedding. It is suitable for interactive applications where quick loading is prioritized over maximum raw fidelity.
Publishing
Once your audio segments are merged, you can distribute them to audiobook platforms. Most services accept standard MP3 files, which are ideal for distribution. Ensure your final file meets the platform’s specific technical requirements before uploading.
Upload to platforms
Export your final audio as an MP3 file. This format is widely supported by audiobook distributors and ensures compatibility across most devices and apps. Upload the file directly to your chosen distributor’s portal.
Metadata setup
Add accurate metadata including title, author, narrator, and chapter titles. Correct metadata ensures your audiobook appears correctly in search results and library catalogs. Double-check all details before submission to avoid rejections or delays.
Quality check
Listen to the entire audio file before publishing. Check for consistent pacing, volume levels, and any remaining artifacts. Ensure the voice tone matches your manuscript’s genre. A final review helps prevent listener complaints and returns.

Cost Management
Generating audiobooks requires careful budget planning since costs scale directly with text volume. Understanding the pricing structure helps you estimate total production expenses before starting your project, ensuring you avoid unexpected charges during the writing process.
Tracking your usage allows you to stay within budget limits and plan future titles efficiently. Regular monitoring prevents running out of credit mid-sentence, keeping your publishing workflow smooth and predictable.
Estimate costs
Pricing is straightforward: $10 per 1 million characters, or $0.01 per 1,000 characters. Spaces and punctuation count toward the total, so ensure your manuscript is clean before calculating. Failed requests are free, so only successful generations incur charges.
For a typical 50,000-character novel, expect to spend approximately $0.50. This low cost makes self-publishing accessible without large upfront investments in voice talent or studio time.
- $0.01 per 1,000 characters
- Spaces and punctuation count
- Failed requests are free
Monitor usage
Your account dashboard displays real-time balance, characters used, and remaining trial credit. Check this page regularly to track consumption across multiple chapters or series.
The system keeps request metadata for billing purposes, allowing you to verify exactly which texts were processed and charged.
Top up balance
Add funds via USDT on the TRON network (TRC-20) with a $5 minimum. Credits appear automatically within minutes, ensuring no downtime in your production pipeline.
Balance never expires, so you can top up once and use it across many projects. New keys replace old ones instantly, keeping your account secure and active.
- USDT (TRC-20) only
- $5 minimum top-up
- Automatic credit within minutes
- No expiration
Troubleshooting
Even with a clean manuscript, technical hiccups can interrupt your audiobook workflow. Understanding common API behaviors and configuration limits helps you resolve issues quickly without losing progress or balance.
Common errors
Requests fail if the input text exceeds 4096 characters or if the API key is invalid. Failed requests incur no charge, so you can retry safely. If the queue is busy, synthesis may take longer than usual because requests are processed sequentially.
- Ensure text length is under 4096 characters per request.
- Verify your Bearer token is active and correct.
- Check that your balance covers the character count.
Speed adjustments
Adjust the speed parameter between 0.25 and 4.0 to fit narration pacing. Increasing speed reduces the audio duration but does not change the character count or cost. Use lower speeds for complex passages to improve clarity.
- Speed affects time, not character count.
- Range is 0.25 to 4.0.
- Adjust based on listener comfort.
Language issues
The API guesses the language from the script, defaulting to English for Latin text. For non-English Latin-script text, always set the language code explicitly. This prevents mispronunciation and ensures the voice matches your target audience.
- Set language code for non-English Latin text.
- Default guess may be incorrect.
- Supported languages include 23 options.
Voice Consistency
Consistent narration is critical for audiobooks. Using the same voice preset throughout your manuscript prevents jarring transitions that distract listeners. Our API supports eight distinct presets, each designed for specific narrative styles, ensuring your audio feels cohesive from start to finish.
Stick to one voice
Select a single preset like 'aria' for warm narration or 'atlas' for neutral news-style reading. Do not switch between voices unless the character changes explicitly. Each voice has a unique timbre; mixing them creates an inconsistent listening experience.
- Choose one preset per book.
- Avoid alternating between male and female voices randomly.
- Define your narrator before starting synthesis.
Avoid switching
Changing voices mid-chapter can confuse your audience. If you need a different tone for a specific section, stick to the same voice but adjust the speed parameter (0.25–4.0) to vary the delivery pace slightly without altering the voice identity.
- Use speed adjustments for pacing.
- Keep voice ID constant.
- Plan voice usage in your outline.
Maintain tone
Ensure the chosen voice matches your genre. A bright voice like 'nova' might feel too upbeat for a serious mystery, while 'orion' provides a deep, documentary tone suitable for non-fiction. Test short samples to verify the emotional fit before generating the entire book.
- Match voice to genre mood.
- Test samples before bulk generation.
- Review tone consistency across chapters.
Legal Considerations
When you generate audio using our API, you receive the file directly. The facts confirm that input text is used only for synthesis and not for training, and generated audio is not stored on our servers. This gives you immediate possession of the output for your projects.
Commercial rights
You can use the generated audio files for commercial purposes, such as publishing audiobooks or selling digital content. However, our product facts do not explicitly grant broader intellectual property rights or a standalone commercial license for the voice engine itself. You own the specific audio file you generated, but you do not acquire rights to the underlying voice model.
- You can sell the audio files.
- You can embed audio in paid courses.
- No broader IP rights to the voice model are granted.
Watermark presence
Every audio file generated by our system carries an inaudible watermark that marks it as synthetic speech. This watermark is embedded automatically during synthesis and cannot be removed. It serves as a marker of origin but does not affect the audible quality or usability of your audiobook.
Attribution
While attribution is not strictly required by our license terms, it is good practice to note that the audio was generated by AudioPage. Since the voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and LibriVox public-domain readings, acknowledging the source adds transparency for your listeners.

Final Review
Before publishing, listen to generated samples to verify audio quality and voice consistency. Ensure the selected preset matches your manuscript’s tone and that the speech remains clear throughout.
Listen to samples
Play short segments to catch any unnatural pauses or artifacts. Verify that the watermark is present but inaudible, confirming the synthetic nature of the audio.
- Test different voice presets to find the best fit
- Check for clarity in complex sentences
Check pacing
Review the speed settings (0.25–4.0) to ensure the narration flows naturally. Adjust the speed if the text feels too rushed or slow for the intended audience.
Submit for review
Once satisfied with the audio quality and pacing, prepare your files for distribution. Ensure all metadata is correct before uploading to your chosen audiobook platform.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Turn your written words into immersive audio stories without writing a single line of code today
23 languages
Any voice speaks any of these languages.
