Choosing Voices for Animation: Presets vs. Cloning

By Voxyto editorial team · Updated 2026-11-19

This guide helps animation startups choose the right voice strategy for their cartoons. We compare our eight licensed preset voices against the concept of custom voice cloning, focusing on copyright safety and API simplicity. You will learn how to select voices that fit your characters without legal headaches or complex setup.

If you are building a series or short-form content, this page gives you concrete facts on pricing, language support, and technical limits. We explain why starting with preset voices often saves time and money, while clarifying what cloning actually involves. Read on to make an informed decision for your production pipeline.

Choosing Voices for Animation: Presets vs. Cloning

Frequently asked questions

What are preset voices?

Preset voices are eight distinct styles based on licensed recordings: Aria, Nova, Luna, Sage, Felix, Milo, Atlas, and Orion. They cover warm narrators, upbeat promos, and deep documentaries. You select one via the voice parameter in your API request. Each voice can speak any of the 23 supported languages, allowing consistent character voices across your animation projects without additional setup.

Are preset voices copyrighted?

The voices are based on LibriTTS-R, which uses recordings from LibriVox public-domain audiobook readings. This means the underlying recordings are in the public domain. However, every generated audio file carries an inaudible watermark marking it as synthetic speech. This helps identify the content as AI-generated while leveraging public-domain source material for your cartoons.

Can I use preset voices in commercials?

Yes, you can use the generated audio in commercials. The voices are derived from public-domain LibriVox readings. Since the source material is public domain, you generally do not face copyright restrictions for the voice itself. The generated files include a synthetic watermark, but this does not restrict commercial use of the audio in your advertising content.

What is LibriVox licensing?

LibriVox provides public-domain audiobook recordings read by volunteers. Our voices are built from these recordings, specifically the LibriTTS-R dataset. This means the original audio sources are free from copyright claims, allowing you to use the resulting synthetic voices in your projects without needing to license individual voice rights from specific actors or studios.

How do I choose a voice for my cartoon?

Select a voice that matches your character’s personality. Use Aria for warm narration, Nova for upbeat energy, or Felix for energetic promos. For neutral news-style delivery, try Atlas. For deep, documentary tones, choose Orion. You can test different voices via the API by setting the voice parameter to the desired preset name in your request.

Do preset voices have accents?

The voices are designed to be clear and standard. While they can speak 23 languages, they do not have strong regional accents unless specified by the language setting. If you send Latin-script text without a language code, it defaults to English. To ensure correct pronunciation for non-English text, explicitly set the language parameter to the ISO code.

Is voice cloning more realistic?

We do not offer custom voice cloning for customers. You cannot upload your own voice or design a new one. Our product provides eight preset voices based on licensed recordings. These presets are realistic within their defined styles but are not unique to your brand. If you need a specific cloned voice, you must use one of the eight available presets or source your own.

What is the cost of cloning?

Since we do not offer voice cloning, there is no specific cloning cost. You pay $10 per 1 million characters for any voice generation, regardless of which preset you choose. All eight presets are included in this standard rate. There are no extra fees for using specific voices or generating longer texts within the character limit.

Can I mix presets and cloning?

You can mix presets in your animations, but you cannot mix them with cloning because we do not offer cloning. You can switch between any of the eight presets (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion) across different API requests. Each request generates audio using one of these eight licensed voice models based on public-domain recordings.

Do voices sound robotic?

The voices are based on real human recordings from LibriVox, so they sound natural and human-like. They are not robotic. However, they are synthetic and carry an inaudible watermark. The quality is consistent with licensed audiobook readings, providing clear, expressive speech suitable for cartoons without the uncanny valley effect of older TTS systems.

What audio formats are best for cartoons?

You can choose from mp3, wav, opus, flac, or pcm. For cartoons, mp3 or wav are common choices. Mp3 offers smaller file sizes for web delivery, while wav provides uncompressed quality. You specify the format using the response_format parameter in your API request. Select the format that best fits your animation pipeline’s storage and playback requirements.

Can I change the speed of the voices?

Yes, you can adjust the speed between 0.25 and 4.0. This allows you to make voices sound faster for energetic scenes or slower for dramatic effect. Set the speed parameter in your API request. The voices remain clear at extreme speeds, ensuring your cartoon characters remain understandable regardless of the pacing you choose.

Is there a limit to how much I can generate?

Each API request accepts up to 4096 characters of text. For longer scripts, you must split the text into parts and join the audio files. There is no hard limit on total generation, only on request size. You can generate as much content as your balance allows, with each request processed sequentially, potentially queuing under load.

Do I need to attribute the voices?

The voices are based on public-domain LibriVox recordings, so attribution is not strictly required by copyright law. However, each generated file contains an inaudible watermark identifying it as synthetic speech. You may choose to credit Voxyto or the source dataset, but it is not a mandatory requirement for using the audio in your animation projects.

Understanding Voice Options

Sound wave illustration for: Understanding Voice Options

Animation startups face a critical choice: use our eight licensed preset voices or pursue custom cloning. Presets offer immediate consistency and copyright safety, leveraging LibriVox recordings for reliable character voices without legal headaches. Cloning, while powerful, introduces complexity in rights management and technical setup.

Preset Voices Overview

Our API provides eight distinct voice presets, ranging from warm narrators to deep documentary tones. Each voice is built from licensed LibriVox recordings, ensuring clear commercial rights. This approach allows you to generate consistent character voices directly via a simple POST request, avoiding the need for complex voice design or licensing negotiations.

  • Eight distinct voice personalities
  • Built on licensed LibriVox data
  • Immediate commercial rights

Custom Voice Cloning Overview

Custom cloning involves training a model on specific audio samples to replicate a unique voice. While this can create highly personalized characters, it requires significant effort to source rights, train the model, and manage the resulting assets. It is not a built-in feature of our current API.

Key Differences

Presets offer speed and simplicity with zero training time, whereas cloning demands dedicated effort and rights verification. For most animation startups, presets provide a balance of quality, cost, and ease of integration. Cloning is better suited for specific, high-value characters where a unique sound is essential.

Voice Characteristics and Types

Sound wave illustration for: Voice Characteristics and Types

Voxyto provides eight distinct voice presets designed for animation, split evenly between female and male options. Each voice is derived from licensed LibriVox recordings, ensuring a consistent, copyright-safe foundation for your cartoon projects without the legal overhead of traditional voice acting.

Female Voice Options

Aria offers a warm, clear narration style ideal for gentle storybooks. Nova brings a bright, upbeat energy perfect for lively advertisements. Luna provides a soft, storytelling tone for intimate scenes, while Sage delivers a calm, instructional voice suitable for educational content.

  • Aria: Warm, clear narrator
  • Nova: Bright, upbeat
  • Luna: Soft, storytelling
  • Sage: Calm, instructional

Male Voice Options

Felix injects energetic enthusiasm, making him perfect for promotional clips. Milo offers a friendly, conversational tone for character dialogues. Atlas provides a neutral, news-like delivery, while Orion delivers a deep, documentary-style voice for serious or epic narratives.

  • Felix: Energetic, promo
  • Milo: Friendly, conversational
  • Atlas: Neutral, news
  • Orion: Deep, documentary

Voice Personality Traits

Each preset carries a specific personality trait that helps define character consistency across episodes. These traits are baked into the synthesized audio, eliminating the need for complex emotion tags. Simply select the voice that matches your character's role and let the API handle the delivery.

Licensing and Copyright

Sound wave illustration for: Licensing and Copyright

Our voices are built from licensed LibriVox recordings, giving animation startups a clear legal foundation. This approach removes the uncertainty often associated with sourcing narration for cartoons, allowing teams to focus on storytelling without worrying about obscure copyright traps.

By using pre-licensed public-domain and CC BY 4.0 sources, we handle the complex rights management for you. The generated audio carries an inaudible watermark, marking it as synthetic speech while maintaining the clarity and legal safety needed for commercial distribution.

LibriVox Licensing Explained

LibriVox recordings are generally in the public domain, while some newer contributions use CC BY 4.0 licenses. This mix ensures broad usability for commercial projects without requiring individual permissions for each voice preset.

  • Public domain recordings have no expiration or attribution requirements.
  • CC BY 4.0 allows commercial use with proper attribution.

Commercial Use Permissions

All preset voices are cleared for commercial use in animations, games, and podcasts. You do not need to pay additional royalties or seek permission from the original readers, streamlining your production workflow.

Cloning Rights Issues

Since we do not offer custom voice cloning, you avoid the legal gray areas of mimicking specific celebrities or actors. There is no risk of unauthorized likeness usage, as our voices are distinct presets based on licensed recordings.

Cost Comparison

Sound wave illustration for: Cost Comparison

Choosing the right voice strategy directly impacts your animation startup’s bottom line. With preset voices, costs are transparent and predictable: you pay $0.01 per 1,000 characters, with no minimum spend or monthly fees. Failed requests are free, ensuring you only pay for successful audio generation.

Preset Voice Costs

Preset voices are included in the standard API pricing. You generate audio using one of eight licensed voices for the same rate, regardless of the character’s personality. This simplicity eliminates hidden fees, making budgeting straightforward for short episodes or pilot projects.

Cloning Costs and Effort

While we do not offer custom voice cloning, the effort to achieve consistency lies in your selection process. You must choose the best preset that matches your character’s tone, potentially requiring multiple test generations to find the perfect match among the eight options.

Long-Term Budget Impact

For long-form series, there are no volume discounts, but the lack of cloning infrastructure fees keeps costs linear. You can start with the free trial credit of $0.50 and scale up via USDT top-ups as your production grows, without committing to expensive enterprise contracts.

Technical Implementation

Sound wave illustration for: Technical Implementation

Integration is straightforward: send a POST request to our API with your script and desired voice. The system returns a complete audio file, ready for immediate download and insertion into your animation timeline. This eliminates the need for complex streaming infrastructure or handling partial audio chunks during playback, keeping your production pipeline simple and reliable.

Unlike custom voice cloning, which often requires training models on specific voice samples and managing larger compute resources, our preset voices are instantly available. You simply select from eight distinct character types and start generating. There is no waiting period for model training, allowing you to iterate on dialogue and scene pacing rapidly without technical bottlenecks.

API Integration Simplicity

Our endpoint follows the standard speech API pattern, making it easy to use with existing libraries. You send text and voice selection, and receive a high-quality audio file. This simplicity means your developers can integrate voice generation without building custom infrastructure for audio handling or streaming.

The lack of required streaming protocols reduces complexity in your animation software. You receive the full voiceover asset to download and integrate into your animation software.

  • Standard JSON request format
  • Immediate audio file return
  • No streaming infrastructure needed
  • Works with standard HTTP clients

Cloning Setup Complexity

While we do not offer custom cloning, understanding the alternative helps. Traditional cloning requires collecting voice samples, training models, and often managing additional storage and compute costs. Our preset approach removes this layer entirely.

You avoid the technical debt of maintaining voice models. Each generated file is self-contained, ensuring consistent quality without network interruptions during playback when you decide to implement future streaming needs.

  • No training data collection
  • No model management
  • No additional compute costs

Development Time

With presets, you can go from script to audio in minutes. This speed allows animation startups to prototype scenes quickly and adjust dialogue without waiting for long rendering or training processes.

The straightforward API reduces development time significantly. Your team can focus on storytelling and visual design rather than debugging audio infrastructure or managing complex voice synthesis parameters.

  • Minutes from text to audio
  • No model training wait
  • Quick iteration cycles
  • Minimal code changes required

Quality and Consistency

Sound wave illustration for: Quality and Consistency

Consistency is critical for animation branding. Our eight preset voices—like Aria’s warm narration or Felix’s energetic promo style—deliver uniform character audio across episodes, ensuring your cartoon’s auditory identity remains stable without complex management.

Speech sounds natural and clear, generated from licensed LibriVox recordings. The API processes requests sequentially, so while short phrases take a few seconds, longer texts generate at roughly real-time speed, maintaining audio quality even under load.

Voice Consistency Across Episodes

Each voice is a fixed preset, meaning every request using "nova" or "atlas" produces the exact same tonal quality. This eliminates the variability often found in human performers, allowing you to generate hundreds of lines without worrying about subtle shifts in tone or pitch between recording sessions.

  • Fixed presets ensure identical tone across thousands of requests.
  • No drift over time or between different recording sessions.
  • Easy to maintain character identity across long series.

Naturalness of Speech

Voices are built from licensed LibriTTS-R recordings, providing a high baseline of human-like cadence. The system guesses language from script, defaulting to English for Latin text, but you can explicitly set the language parameter for non-English Latin-script text to ensure accurate pronunciation and flow.

  • Based on licensed LibriVox public-domain audiobook readings.
  • Inaudible watermark marks synthetic speech clearly.
  • Natural cadence without robotic pauses.

Emotional Range

While we do not offer emotion tags or custom voice cloning, each preset carries a distinct personality. Sage offers calm, instructional tones, while Orion provides a deep, documentary feel. You select the appropriate preset for the scene’s mood rather than adjusting emotional parameters within a single voice.

  • Distinct personalities for each of the 8 voices.
  • No emotion tags; choose the right preset for the mood.
  • No custom cloning available to modify emotional nuance.

Use Case Scenarios

Sound wave illustration for: Use Case Scenarios

For animation startups, selecting the right voice workflow hinges on content scale and budget. Preset voices offer immediate consistency across short clips, while cloning introduces complexity for long-form narratives. Understanding these trade-offs helps teams allocate resources efficiently without over-engineering simple projects.

Short Form Content

Short animations like ads or social media clips benefit from preset voices. You can generate audio instantly using the API’s 4096-character limit. With no cloning setup required, you maintain consistent branding across episodes without waiting for model training.

  • Ideal for promo videos under 60 seconds
  • Instant generation with zero setup time
  • Cost-effective for high-volume, low-complexity content

Long Form Series

Series with recurring characters need voice consistency. While presets are available, cloning ensures unique character identity. However, cloning requires effort and may not be offered as a custom service here. Use presets like 'sage' for instructional content or 'nova' for upbeat episodes to maintain tone across seasons.

  • Maintains character identity across episodes
  • Requires careful voice selection for consistency
  • Higher long-term consistency than ad-hoc presets

Interactive Media

Games and apps often require varied voice lines. Presets allow rapid iteration. You can switch voices for different characters or emotions using the 8 available presets. The API’s speed supports near real-time generation, keeping user experience smooth without waiting for long synthesis jobs.

  • Rapid iteration for game dialogue
  • Switch voices easily via API parameters
  • Supports dynamic content updates
Use Case Scenarios

Language Support

Sound wave illustration for: Language Support

Voxyto supports 23 languages, covering major global markets for animation. You can specify any of these languages explicitly or let the API guess based on the text script. This flexibility ensures your characters sound authentic whether you are producing content for English, Japanese, or Hindi audiences.

The system handles non-Latin scripts correctly, but for Latin-script languages like Dutch or Norwegian, you should set the language field to avoid defaulting to English. This small detail prevents unwanted accent shifts in your final audio files.

Available Languages

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. Any voice can speak any of these languages, giving you vast creative freedom without needing multiple voice actors.

  • 23 supported languages
  • Any voice works in any language
  • Script-based language guessing

Multi-Language Projects

You can generate audio in different languages using the same voice presets. This is ideal for dubbing or creating multilingual versions of your cartoon episodes. Just change the language parameter in your API request to match the target language.

Accent Variations

While we do not offer custom cloning, the eight preset voices provide distinct personalities that can mimic various accents naturally. For example, aria offers a warm, clear narrator tone suitable for many English dialects, while felix brings an energetic promo style. Remember that if language is not set, Latin text defaults to English, so always specify the language for non-English texts to ensure correct pronunciation.

Audio Formats and Delivery

Sound wave illustration for: Audio Formats and Delivery

Our API returns audio in MP3, WAV, OPUS, FLAC, or PCM formats, giving animation studios the flexibility to choose the right balance between file size and quality. Each format serves a different part of your production pipeline, from quick previews to final broadcast master files.

Supported Formats

MP3 is ideal for web delivery and quick reviews due to its small size. WAV and FLAC offer uncompressed or lossless quality for editing. OPUS provides efficient compression for streaming, while PCM offers raw data for direct processing.

  • MP3: Best for web and previews
  • WAV/FLAC: Uncompressed/lossless for editing
  • OPUS: Efficient compression
  • PCM: Raw data for processing

File Size and Quality

Higher bitrates in WAV and FLAC preserve more detail, crucial for complex soundscapes in animation. MP3 and OPUS reduce file sizes significantly without noticeable loss for most cartoon applications, keeping your storage costs low.

Streaming vs. Download

The API delivers complete audio files via POST requests, not streaming audio chunks. This means you receive the full voiceover asset to download and integrate into your animation software, ensuring consistent quality without network interruptions during playback.

Decision Framework

Sound wave illustration for: Decision Framework

Choosing the right voice strategy depends on your production scale and creative needs. Presets offer immediate consistency with zero setup time, while custom cloning requires effort to capture a unique sound. Understanding these trade-offs helps animation startups manage budgets and timelines effectively without over-engineering their audio pipeline.

Hybrid approaches often provide the best balance, using presets for background characters to save time and custom cloning for protagonists to ensure distinct branding. This section helps you decide which path aligns with your specific project requirements and long-term goals.

When to Choose Presets

Opt for presets if you need consistent, copyright-safe voices quickly. This is ideal for short-form content or series with many characters where unique vocal signatures are less critical. Preslets like Aria or Felix provide reliable, high-quality narration without additional development effort.

When to Choose Cloning

Opt for custom cloning if your brand requires a distinct, ownable voice that cannot be replicated by existing presets. This suits projects where character identity is paramount, though it requires more effort to record and train the voice model.

Hybrid Approaches

Combine both methods by using presets for background characters and custom cloning for protagonists. This strategy maximizes creative impact for main roles while keeping production costs and time manageable for supporting roles.

- Presets for background characters - Custom cloning for protagonists - Balanced cost and creativity - Efficient production workflow

Common Pitfalls

Sound wave illustration for: Common Pitfalls

Animation startups often assume that because their voices are licensed, they are free from all copyright risks. However, the distinction between the underlying recording license and the generated synthetic output can be tricky. While our voices are built from LibriVox and LibriTTS-R recordings, the synthetic nature of the output means you must ensure your specific use case aligns with the commercial permissions granted by those source licenses.

Another common issue is expecting perfect consistency without managing character profiles. Since we do not offer custom voice cloning, relying on different preset voices for the same character across episodes can break immersion. You must manually select and stick to one preset (e.g., always using "sage" for your narrator) to maintain auditory continuity.

Copyright Infringement Risks

While our voices are derived from licensed public-domain or CC BY 4.0 recordings, the synthetic watermark marks the file as AI-generated. Ensure your distribution platform accepts synthetic audio. Some broadcasters or streaming services have specific clauses regarding AI content that you must review before publishing.

  • Check source license (LibriVox vs. LibriTTS-R) for commercial use.
  • Verify platform policies on watermarked synthetic audio.
  • Do not claim human voice actors if using presets.

Inconsistent Voice Quality

Without custom cloning, voice consistency relies entirely on your choice of preset. If you switch from "nova" to "luna" between episodes, the character’s tone shifts dramatically. Always document which preset corresponds to each character in your production pipeline to avoid accidental casting changes.

  • Lock one preset per character.
  • Use the same speed setting for consistency.
  • Avoid switching voices mid-scene.

Unexpected Costs

Characters exceeding 4096 characters per request will be truncated, leading to incomplete audio files that still incur charges. Additionally, failed requests are free, but successful ones are billed immediately. Ensure your input text is pre-chunked correctly to avoid wasting credits on partial outputs.

  • Split text into 4096-character chunks.
  • Check balance before large batches.
  • Failed requests are refunded, but successful ones are not.

Final Recommendations

Sound wave illustration for: Final Recommendations

For animation startups, the most efficient path is leveraging our eight licensed presets. You get consistent, copyright-safe voices immediately without legal overhead. Use the API to generate dialogue in batches, keeping each request under 4096 characters to ensure smooth processing. This approach lets your team focus on storytelling rather than voice engineering.

Stick to the standard formats like MP3 or WAV for easy integration. Since voices are based on LibriVox recordings, you avoid trademark issues entirely. This simplicity reduces your time-to-market significantly, allowing you to prototype episodes faster than using traditional voice cloning pipelines.

Best Practices for Startups

Start with a single voice type for your main character to maintain brand consistency. Use the free trial credit to test different presets before committing to a budget. Keep your text concise to stay within the 4096-character limit per request, ensuring faster generation times.

Document your voice choices early. Since each account has one active API key, keep your documentation clear about which voices are used for which characters. This helps when scaling later or handing off projects to new developers.

  • Use tts-1-hd for high-quality main characters.
  • Split long scripts into chunks under 4096 characters.
  • Test voices with short samples before full production.

Scaling Your Production

As your series grows, rely on the 60 requests per minute limit to batch generate episodes. If you need more speed, issue a new API key to refresh your rate limit window. Remember that requests are processed sequentially, so plan your generation schedule to avoid queues during peak times.

For multi-language projects, use the language parameter to specify ISO codes. This ensures accurate pronunciation across the 23 supported languages. You can reuse the same voice presets across different languages, maintaining character identity while expanding your global reach.

  • Batch generate audio for entire episodes.
  • Issue new keys if you hit rate limits.
  • Specify language codes for accurate pronunciation.

Future-Proofing Your Audio

Our API returns audio files directly, so you own the output immediately. There is no dependency on our servers for playback. Store your generated audio in your own cloud storage to ensure long-term access. Since the voices are based on public domain recordings, your rights remain clear even if the platform changes.

Avoid relying on streaming features since we do not offer real-time WebSocket audio. Download the full audio files for your production pipeline. This ensures you have complete control over your assets, ready for editing, dubbing, or distribution without technical constraints.

  • Store audio files in your own cloud.
  • Download full files for editing stability.
  • No dependency on live API for playback.

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. You receive the full voiceover asset to download and integrate into your animation software.

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. Python documentation
  3. LibriTTS-R paper (arXiv 2305.18802)
  4. Tether (USDT)
  5. LibriTTS-R corpus (OpenSLR 141)
  6. Creative Commons Attribution 4.0
  7. Node.js documentation
  8. MDN: Web Speech API
  9. openai-node on GitHub
  10. TRON developer hub: TRC-20
  11. IETF RFC 7845: Ogg Opus
  12. curl documentation

More on this site