Text-to-Speech API for Audiobook Creators

By ReadCast editorial team · Updated 2026-11-19

This guide evaluates ReadCast API for independent authors and audiobook narrators seeking consistent, character-based synthesis without enterprise contracts. We analyze voice quality, language support, and pricing transparency to help you decide if this tool fits your production pipeline.

Readers will find concrete data on generation speed, format options, and technical limitations. Whether you are narrating fiction or educational content, this review provides the specific metrics needed to integrate ReadCast API into your workflow efficiently.

Text-to-Speech API for Audiobook Creators

Frequently asked questions

Which voice is best for fiction?

Aria offers a warm, clear narrator style, while Luna provides a soft, storytelling tone. Sage is suitable for instructional content, and Orion delivers a deep, documentary feel. You can select any of the eight presets based on your specific narrative needs, as all voices support the full range of 23 languages effectively.

Is this API better than ElevenLabs for audiobooks?

ReadCast API focuses on reliable, character-based pricing at $0.01 per 1,000 characters with no monthly plans. Unlike some competitors, it uses standard OpenAI SDK formats, making integration straightforward. It offers 23 languages and eight distinct voice presets, providing a predictable cost structure for independent authors without enterprise contracts or complex tiering.

Can I use this API for commercial audiobooks?

Yes, the API is designed for content creators, including independent authors and narrators. Input text is used only to produce audio and is not used for training. Generated audio files are returned in the response and are not stored by the service, allowing you to retain full rights to your commercial productions.

How many languages does this API support?

The API supports 23 languages, including English, Spanish, French, German, Japanese, Korean, and Chinese. Any voice can speak any of these languages. If the language parameter is not set, the system guesses from the script, defaulting to English for Latin text, so you should explicitly set it for non-English Latin-script content.

What is the maximum character limit per request?

Each request accepts up to 4,096 characters in the input field. For longer manuscripts, you must split the text into parts of this size and join the resulting audio files. This ensures consistent processing speed and allows you to manage memory usage effectively when generating full-length audiobooks from large texts.

Does this API offer voice cloning?

No, custom voice cloning is not offered. You cannot upload your own voice or design custom voices. The service provides eight preset voices based on licensed recordings from LibriTTS-R and LibriVox. Each generated file carries an inaudible watermark marking it as synthetic speech, but no personal voice replication is available.

Is the audio quality comparable to human narrators?

The voices are based on licensed recordings from LibriTTS-R and LibriVox public-domain audiobook readings. While synthetic, they offer consistent quality without the fatigue of human narration. The audio carries an inaudible watermark indicating it is synthetic speech, ensuring transparency about the origin of the voice while maintaining high production standards for your projects.

How do I handle special characters in my text?

Special characters and spaces count as characters toward your 4,096-character limit per request. The API processes standard text inputs directly. There is no SSML support for custom pronunciation or emphasis, so you must rely on the natural intonation of the selected voice presets. Ensure your text fits within the character constraints for optimal results.

Can I generate audio in real-time?

No streaming or WebSocket support is available for real-time partial audio. Requests are processed sequentially, meaning under load, a request may wait in a queue. Typical speed is roughly real-time on long texts, but short phrases may take a few seconds. You receive the complete audio file after synthesis finishes.

What is the refund policy for failed requests?

Failed requests are free. If synthesis fails, the charge is returned in full. This applies to your account balance or top-up funds. You only pay for successfully generated audio files. This policy ensures you are not charged for errors or processing issues, maintaining a fair cost structure for your audiobook production workflow.

Do I need technical skills to use this API?

Basic technical skills are helpful but not required. You can use the official OpenAI SDKs by changing the base URL, or send standard HTTP requests. The /get-api-key/ page provides a ready curl example and shows your balance. However, understanding JSON bodies and API endpoints simplifies integration into your existing publishing pipeline.

Is there a limit on the number of audio files I can generate?

There is no limit on the number of files you can generate, only on your character count and rate limits. You can make 60 requests per minute with an active key. Longer texts require splitting into chunks. Your balance does not expire, allowing you to generate as much audio as your credits allow.

Can I change the voice after generating the audio?

No, you cannot change the voice after generation. Each request produces audio with a specific voice preset. To change the voice, you must send a new request with the desired voice parameter. The API does not store generated audio, so you must save the files locally if you wish to keep them.

How is the audio watermark implemented?

Every generated file carries an inaudible watermark that marks it as synthetic speech. This watermark is embedded directly into the audio data upon generation. It serves to identify the content as AI-generated without affecting the listening experience. The watermark is present in all output formats, including MP3, WAV, OPUS, FLAC, and PCM.

Criteria for Evaluation

Sound wave illustration for: Criteria for Evaluation

This guide evaluates ReadCast API specifically for audiobook production, focusing on the practical needs of independent authors. We assess voice quality, language support, and pricing transparency to determine suitability for converting manuscripts into audio files.

Voice Quality

Voice quality is the primary metric for audiobook creators. We evaluate the eight available voices for natural flow, distinct personalities, and consistency across long-form content. Each generated file includes an inaudible watermark, ensuring synthetic speech is clearly identified without disrupting the listening experience.

Language Support

Support for 23 languages is critical for global distribution. We check if any voice can speak any supported language and how accurately the API handles non-English Latin-script text without explicit language tags. Languages like Vietnamese or Thai are noted as unsupported.

Pricing Transparency

Pricing is calculated per character, including spaces and punctuation. We verify the cost of $10 per million characters against typical manuscript lengths. The model ensures no hidden fees for failed requests, which are refunded in full, providing predictable budgeting for self-published authors.

Voice Variety and Quality

Sound wave illustration for: Voice Variety and Quality

ReadCast API offers eight distinct voice presets designed to suit various audiobook styles. Each voice provides a consistent tone, from the warm narration of Aria to the deep, documentary style of Orion, allowing creators to match the narrative mood precisely.

The engine generates natural-sounding speech by leveraging licensed recordings, ensuring that the output feels professional and consistent across long-form content without the variability of random selections.

Distinct Personalities

The eight available voices range from female presets like Luna’s soft storytelling style to male options like Felix’s energetic promo tone. Each voice is based on licensed recordings from LibriVox, ensuring a grounded, human-like quality that suits independent publishing needs.

Creators can select the perfect narrator for their manuscript, whether it requires the calm instructional tone of Sage or the bright, upbeat delivery of Nova. This variety allows for dynamic audiobook production without needing multiple voice actors.

  • 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

Natural Flow

The API processes requests sequentially, aiming for roughly real-time generation speed. Short phrases may take a few seconds to synthesize, while longer texts are split into 4096-character segments to maintain consistent audio quality.

Each generated file includes an inaudible watermark, marking it as synthetic speech. This ensures authenticity without detracting from the listening experience, providing a clean audio stream for distribution.

Emotional Range

While the voices do not support dynamic emotion tags or SSML controls, their inherent tonal qualities provide a baseline emotional context. For example, Orion’s deep tone suits serious documentaries, while Luna’s softness enhances fairy tales.

For nuanced emotional shifts, creators can switch voices between chapters or scenes. This manual control offers more flexibility than fixed emotion tags, allowing precise curation of the audiobook’s emotional arc.

Language Coverage

Sound wave illustration for: Language Coverage

ReadCast API supports 23 languages, allowing creators to produce multilingual audiobooks from a single account. The system includes major markets like English, Spanish, German, French, Japanese, and Arabic, covering a broad range of global audiences without requiring separate integrations.

23 Languages Supported

The platform handles 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, providing flexibility for diverse narration styles.

  • Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish

Cross-Linguistic Accuracy

Voices are based on licensed recordings from LibriTTS-R and LibriVox public-domain audiobook readings. The API guesses the language from the script, defaulting to English for Latin text, though setting the language parameter is recommended for non-English Latin-script content to ensure accuracy.

Dialect Considerations

While the API supports 23 languages, it does not offer specific dialect variations for each language. Creators should test samples to ensure the standard voice aligns with their target regional audience, as no custom pronunciation lexicons or SSML are available to fine-tune dialects.

Ease of Integration

Sound wave illustration for: Ease of Integration

The API mirrors the OpenAI speech endpoint structure, allowing seamless integration for teams already familiar with standard TTS workflows. By simply adjusting the base_url to https://api.link02.info/v1 and supplying your unique API key, existing OpenAI SDKs function without modification.

OpenAI SDK Compatibility

Standard libraries handle authentication and request formatting automatically. This reduces boilerplate code and accelerates prototyping for indie developers who need reliable audio generation without learning a new proprietary protocol.

Documentation Quality

Clear endpoint definitions for /v1/audio/speech and /v1/voices are provided. Examples demonstrate correct JSON bodies, including required fields like model and voice, ensuring accurate request construction from the first attempt.

Developer Experience

Integration is straightforward: send text via POST, receive audio files. The consistent request shape supports automated pipelines for converting manuscript chapters into audio assets efficiently.

  • Use standard HTTP clients or OpenAI SDKs.
  • Supports tts-1 and tts-1-hd models.
  • Simple JSON request body structure.

Pricing Models Compared

Sound wave illustration for: Pricing Models Compared

Understanding the cost structure is critical for audiobook producers managing variable project lengths. ReadCast API uses a straightforward per-character model, eliminating the ambiguity of time-based billing where pause durations can distort costs. This approach ensures you pay only for the text generated, providing predictable expenses for manuscripts of any length.

Unlike traditional subscription services that charge for access rather than usage, our pay-as-you-go model scales with your output. You are not locked into monthly tiers; you simply fund your account and consume credits as needed, making it ideal for creators with irregular publishing schedules.

Per Character vs Per Minute

Audio length varies significantly based on voice choice and speed settings. A minute of audio might contain 150 characters at normal speed or 60 at a slow, dramatic pace. Per-minute billing penalizes slower narrations, whereas per-character billing remains consistent regardless of delivery speed.

The price is $0.01 per 1,000 characters. This includes spaces and punctuation. You can calculate exact costs before generating by counting characters in your manuscript file, ensuring no surprise charges when the final audio duration differs from your initial estimate.

  • Fixed rate: $10 per 1 million characters
  • Includes spaces and punctuation
  • No penalty for slow narration speeds

Subscription vs Pay-as-you-go

There are no monthly plans or seat limits. You do not lose credits if you do not publish in a given month. This contrasts with subscription models that often require minimum monthly commitments regardless of usage volume.

Top-ups are processed via USDT (TRC-20) and credited automatically. The balance never expires, allowing you to accumulate funds during quiet periods and deploy them during high-volume production weeks without administrative overhead.

  • No monthly commitments
  • Credits never expire
  • No per-seat fees

Hidden Costs

Failed requests are free, so you do not pay for errors. The only potential cost is the time spent splitting long texts into 4,096-character chunks, which you can automate via the API.

There are no hidden fees for format selection (MP3, WAV, OPUS, FLAC, PCM) or speed adjustments. The price per character remains identical regardless of the output format or voice preset chosen.

  • Failed requests are free
  • No format or speed fees
  • No VAT or invoice charges

Performance and Speed

Sound wave illustration for: Performance and Speed

ReadCast API processes synthesis requests sequentially, meaning throughput depends on current system load. Understanding these limits helps audiobook creators plan their production pipelines effectively without unexpected delays.

The system handles requests at a steady pace, with typical generation speed matching roughly real-time audio duration. Short phrases may take a few seconds, while longer texts are processed in continuous segments.

Generation Latency

Generation speed averages approximately 25 characters of text per second of audio output. This real-time ratio holds for long texts, though short phrases may require a few additional seconds for processing overhead.

Throughput Limits

Each API key is limited to 60 requests per minute once funded, or 10 requests per minute during the initial trial phase. These limits ensure consistent quality and prevent server overload during high-demand periods.

Queue Management

Requests are processed in the order they are received. Under heavy load, a request may wait in a queue before synthesis begins. This sequential processing ensures that each audio file is generated with consistent quality, though it may introduce minor delays during peak usage times.

Audio Format Options

Sound wave illustration for: Audio Format Options

The API allows creators to choose from five distinct audio formats, balancing file size and fidelity for different distribution channels. Each format serves a specific technical need, from web streaming to archival preservation.

Selecting the right format ensures compatibility with major audiobook platforms while maintaining the integrity of the generated speech.

MP3 Standardization

MP3 is the default and most widely supported format, ideal for web playback and standard audiobook distributors. It offers a strong balance between quality and file size, making it suitable for most commercial releases.

  • Universal player compatibility
  • Optimized file sizes for distribution

Lossless Options

For archival purposes or high-fidelity masters, WAV and FLAC provide lossless quality, while Opus is a lossy codec. FLAC and WAV preserve every detail of the synthetic voice, while Opus offers high efficiency without significant quality loss.

PCM provides raw, uncompressed audio data for direct processing or specialized hardware integration.

  • WAV: Standard uncompressed audio
  • FLAC: Compressed lossless audio
  • Opus: Efficient high-quality codec
  • PCM: Raw audio data

Streaming Compatibility

While the API does not support real-time streaming or WebSocket connections, all generated files can be streamed once fully delivered. MP3 and Opus are optimized for playback in modern browsers and mobile apps.

Choose MP3 or Opus for the best balance of compression and playback smoothness in standard streaming contexts.

  • MP3: Best for broad compatibility
  • Opus: High quality, small size
Audio Format Options

Customer Support and Reliability

Sound wave illustration for: Customer Support and Reliability

ReadCast API relies entirely on email for customer support, maintaining a direct line without phone numbers or live chat agents. This approach keeps overhead low, allowing the team to focus on core infrastructure and API stability rather than staffing multiple communication channels. There is no dedicated account manager or enterprise SLA guaranteeing specific response windows for individual clients.

Reliability is managed through standard cloud operations without public uptime certificates or third-party certifications like SOC 2 or HIPAA. Users should expect standard internet variability, with no guaranteed 99.9% uptime contract, but consistent service delivery for character-based synthesis tasks.

Email Support Only

All inquiries, from billing questions to technical issues, are handled via email. There is no phone support or live chat widget. This ensures that every support request is documented and can be referenced later, which is useful for troubleshooting API key or billing discrepancies.

  • No phone number available
  • No live chat widget
  • Email is the sole support channel

Uptime History

The service does not publish a formal uptime percentage or SLA. Reliability is maintained through standard cloud infrastructure practices, but there is no official guarantee of continuous availability. Users should plan for occasional latency spikes during high-demand periods, especially when processing long texts.

Response Times

Support response times vary based on queue volume. During peak hours, emails may take several hours to receive a reply. There is no priority tier for paid users, so response speed is generally first-come, first-served. For urgent billing issues, checking your dashboard balance is often faster than waiting for an email reply.

Use Case Suitability

Sound wave illustration for: Use Case Suitability

ReadCast API enables indie authors to produce consistent audio files without enterprise contracts. By leveraging eight distinct voice presets, creators can match the tone of their manuscript precisely, ensuring that character voices remain stable across chapters.

The character-based pricing model aligns directly with manuscript length, eliminating hidden fees. This transparency allows narrators to budget accurately for full-length novels or short story collections without worrying about unexpected overage charges.

Fiction Narration

The eight available voices cover a range of tones suitable for various genres. For instance, 'Luna' offers a soft, storytelling quality ideal for fantasy, while 'Felix' provides an energetic tone for promotional or fast-paced thrillers.

Each generated file includes an inaudible watermark, confirming synthetic origin without affecting playback. This feature is particularly useful for tracking distribution and ensuring authenticity in self-published markets.

  • Distinct personalities for varied characters
  • Watermarked audio for authenticity
  • No emotion tags required

Educational Content

Clear, neutral voices like 'Atlas' and 'Sage' are well-suited for instructional material. The ability to specify language codes ensures accurate pronunciation for non-native speakers, enhancing accessibility for diverse student demographics.

Support for 23 languages allows educators to create multilingual resources efficiently. Since voices can speak any supported language, a single voice preset can be reused across different language versions of the same course.

  • Neutral tones for clarity
  • Multilingual support
  • Consistent character usage

Audiobook Length

The 4,096-character limit per request requires splitting longer texts, but the API handles this seamlessly. Users can join audio files programmatically, ensuring continuous playback for full-length novels.

At roughly real-time speed, generating a 10-hour audiobook takes approximately 10 hours of generation time, plus potential queue time. With a rate limit of 60 requests per minute, large batches are processed efficiently without significant delays.

  • Splits large manuscripts
  • Real-time generation speed
  • Scalable for full novels

Technical Limitations

Sound wave illustration for: Technical Limitations

ReadCast API prioritizes simplicity and speed over complex markup, which means you cannot use SSML tags to control pronunciation or pacing. This design choice keeps the integration lightweight but requires manual text preparation for nuanced readings.

Similarly, the service does not offer custom voice cloning, so you are limited to the eight preset voices. Each generated file contains an inaudible synthetic watermark, which may be relevant for large-scale commercial distribution.

No SSML Support

The API does not support Speech Synthesis Markup Language (SSML). You cannot use tags to adjust pauses, emphasis, or pronunciation within the request body. For specific pronunciation needs, you must modify the input text directly using standard spelling or hyphenation before sending it to the endpoint.

No Voice Cloning

ReadCast API offers eight distinct preset voices but does not allow customers to upload audio samples to create custom cloned voices. If you require a unique narrator voice, you must select from the available options: Aria, Nova, Luna, Sage, Felix, Milo, Atlas, or Orion.

Every audio file generated includes an inaudible watermark identifying it as synthetic speech. This feature ensures transparency but means the output is not indistinguishable from human-recorded audio without additional processing.

Character Limits

Each API request is limited to 4,096 characters of input text. For longer manuscripts, you must split the text into chunks, generate audio for each part, and join the resulting files. This process is straightforward but adds a step to your production pipeline.

Characters include spaces and punctuation. Exceeding the limit will result in a request error, so ensure your text is truncated or split appropriately before submission.

Final Verdict

Sound wave illustration for: Final Verdict

ReadCast API offers a pragmatic, character-based solution that strips away enterprise complexity for independent creators. Its predictable pricing and straightforward API make it an ideal tool for authors who value consistency over complex features.

Best for Indie Publishers

Indie authors benefit from the clear $10 per 1 million characters rate and the ability to process manuscripts in standard 4096-character chunks. The lack of monthly subscriptions means you only pay for what you produce, keeping overhead low during the publishing phase.

  • Character-based billing avoids wasted spend on unused minutes
  • No minimum commitment or monthly fees
  • Simple workflow from text to audio file

Best for Budget-Conscious Creators

With a $0.50 free trial and no hidden costs, creators can test quality before committing. The transparent rate of $0.01 per 1,000 characters allows precise budgeting for audiobook projects, ensuring no surprise charges for failed requests.

  • $0.50 trial credit for immediate testing
  • Failed requests are refunded fully
  • USDT TRC-20 top-ups for low overhead

Best for Technical Users

Developers appreciate the OpenAI SDK compatibility and JSON-based endpoint, allowing easy integration into existing pipelines. The clear documentation and predictable queue behavior make it reliable for automated audiobook generation workflows without needing custom integrations.

  • OpenAI SDK compatible with base_url change
  • Standard JSON request/response format
  • Clear rate limits and character counting

How to Get Started

Sound wave illustration for: How to Get Started

Access your API key by signing up via Google or email on the /get-api-key/ page. Your dashboard displays your key, balance, and usage statistics immediately.

Sign Up Process

Register using Google for immediate access or create an account with an email and password. Each account is site-specific and independent.

First Steps

Use the provided curl example to send your first request. Google sign-ups receive $0.50 credit (50,000 characters) valid for seven days. Email accounts start without trial credit.

  • Top up via USDT (TRC-20) for larger volumes; minimum $5.
  • Split texts exceeding 4,096 characters and join the resulting audio files.

Resources Available

The documentation page details the request schema, including model selection and voice presets. Support is available exclusively via email.

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. The ability to specify language codes ensures accurate pronunciation for non-native speakers, enhancing accessibility for diverse student demographics.

Avatar of the voice aria
ariafemale
Avatar of the voice nova
novafemale
Avatar of the voice luna
lunafemale
Avatar of the voice sage
sagefemale
Avatar of the voice felix
felixmale
Avatar of the voice milo
milomale
Avatar of the voice atlas
atlasmale
Avatar of the voice orion
orionmale

23 languages

Any voice speaks any of these languages.

Sources and further reading

  1. Wikipedia
  2. curl documentation
  3. ISO 639 language codes
  4. TRON developer hub: TRC-20
  5. OpenAI API reference: create speech
  6. FFmpeg documentation
  7. W3C WAI: making audio and video accessible
  8. Python documentation
  9. LibriVox public-domain audiobooks
  10. Tether (USDT)
  11. LibriTTS-R corpus (OpenSLR 141)
  12. Node.js documentation

More on this site