Frequently asked questions
Is voice cloning the same as TTS?
No. Our product is a Text-to-Speech API that uses eight fixed preset voices. We do not offer custom voice cloning where you upload a sample to create a unique voice model. You select from existing presets like Aria or Felix, which are based on licensed recordings, rather than generating a new voice from scratch.
Do I need permission to clone a voice?
Since we do not offer custom voice cloning, this question applies to using our preset voices. These voices are based on recordings from LibriVox public-domain audiobooks. You can use the generated audio commercially without additional permission, as the source material is public domain, though the synthetic output carries an inaudible watermark.
Can I clone a celebrity voice?
No. We do not provide custom voice cloning capabilities. You cannot upload a sample of a celebrity’s voice to replicate it. You must use our eight preset voices (e.g., Orion, Nova) which are distinct, licensed recordings. They are not imitations of specific real-world celebrities or politicians.
Is the watermarked audio detectable?
Yes. Every audio file generated by our API carries an inaudible watermark that marks it as synthetic speech. This watermark is embedded in the output to indicate that the audio was created by our TTS engine, helping distinguish it from human-recorded voiceovers in your animation projects.
Are the preset voices copyrighted?
The voices are based on LibriTTS-R recordings (CC BY 4.0) from LibriVox public-domain audiobooks. While the underlying recordings are public domain, the specific synthetic generation is ours. You can use the output commercially, but the voices are fixed presets, not custom clones, and remain distinct from any specific copyrighted character voice.
Can I use cloned voices commercially?
We do not offer voice cloning, so this applies to our preset voices. Yes, you can use the generated audio commercially. The voices are derived from public-domain LibriVox readings. You can integrate them into animations, cartoons, or other media without needing to pay additional licensing fees per use.
How long does cloning take?
Since we do not offer voice cloning, there is no training period. You get immediate results. Text is converted to audio in real-time or near real-time. Longer texts process at roughly 25 characters of text per one second of audio, with requests processed sequentially in a queue.
Is cloning expensive for small studios?
We do not charge for cloning since we don’t offer it. Our pricing is $10 per 1 million characters. There are no monthly plans or minimum spends. You only pay for what you generate, making it cost-effective for small studios producing animation drafts without hiring voice actors for every take.
Do platforms require labeling cloned audio?
Our audio files include an inaudible watermark identifying them as synthetic speech. While we don’t clone voices, this watermark helps identify the content as AI-generated. Some platforms may require labeling for synthetic media; the watermark serves as a technical indicator, but you should check specific platform guidelines for your distribution needs.
Can I edit cloned voices?
We do not offer voice cloning, so you cannot edit a cloned voice model. However, you can adjust the speed of our preset voices between 0.25x and 4.0x. You can also split long texts into parts to control pacing. The voices themselves are fixed presets and cannot be modified individually.
Is my voice data stored during cloning?
Since we do not offer custom voice cloning, no voice sample of yours is stored. For standard TTS, your input text is used only to produce the audio and is not used for training. Generated audio is returned in the response and not stored. Only request metadata is kept for billing.
What is CC BY 4.0 license?
CC BY 4.0 is a Creative Commons Attribution license. Our preset voices are based on LibriVox public-domain readings licensed under CC BY 4.0. This allows you to use the underlying voice characteristics freely, provided you attribute the source if required by specific platform rules.
Are there legal risks in cloning?
We do not offer voice cloning, so risks associated with custom cloning (like likeness rights) are not applicable. Using our preset voices, which are based on public-domain LibriVox recordings, carries minimal legal risk. You are using synthetic speech from fixed presets, not replicating a specific person’s unique voice identity through cloning technology.
Can I use presets for final broadcast?
Yes. Our eight preset voices (e.g., Aria, Felix, Atlas) are designed for professional use. They are based on licensed recordings and can be used in final broadcast animations. The audio is delivered in standard formats like mp3, wav, opus, flac, or pcm, suitable for post-production pipelines without requiring additional voice actor sessions for every draft.
Defining Voice Cloning
Voice cloning creates a unique digital voice profile by analyzing hours of specific speech samples. The AI learns the distinct tone, pitch, rhythm, and emotional cadence of a target speaker. This allows the model to generate new audio that sounds indistinguishable from the original person, even for dialogue they never previously recorded.
Custom Voice Creation
This process requires dedicated training data to capture nuances like tone, pitch, and rhythm. It typically involves a separate training phase before generation begins. The resulting model is tied to that specific voice identity, making it a distinct product from standard preset voices. Agencies often use this for hero characters requiring a bespoke sound.
Consent and Licensing
Voice cloning often requires explicit consent from the original speaker, especially when their unique vocal timbre is replicated for commercial use. Without clear permission, studios risk legal disputes over who owns the resulting voice asset.
Licensing terms vary significantly. Some platforms grant broad rights, while others restrict usage to specific projects or timeframes. Animation agencies must verify that their license covers all intended distributions, including sequels and merchandise.
Ethical Considerations
When using cloned voices, agencies should ensure the source material was recorded with permission. Public domain recordings, like those from LibriVox, generally allow commercial use, but custom cloning of living talent requires a signed agreement.
Transparency helps avoid reputational risk. Audiences may react negatively if a cloned voice is used deceptively. Clearly attributing the voice source maintains trust with viewers and collaborators alike.
- Obtain written consent for living talent.
- Verify public domain status for historical recordings.
- Define usage rights in client contracts.
Our Approach: Preset Voices
We skip the complexity of voice cloning entirely. Instead of training a unique model on your talent, we provide eight consistent, high-quality preset voices. This allows animation agencies to prototype characters rapidly without the long turnaround times associated with custom cloning projects.
Our approach prioritizes speed and cost-efficiency for draft iterations. You can swap between eight distinct vocal personalities instantly, keeping your production pipeline moving without waiting for custom voice models to be built and licensed.
Voice Cloning Not Offered
ToonVoce does not offer custom voice cloning for customers. We do not upload your own voice recordings to create a digital twin, nor do we provide voice design services. If you need a specific celebrity or actor’s voice, you must license that talent separately.
We focus on eight pre-licensed presets derived from LibriTTS-R and LibriVox readings. These voices are consistent across 23 languages and are ready for immediate use via our API. This eliminates the need for additional legal clearance for voice likeness, streamlining your workflow for quick drafts and storyboards.
- No custom voice cloning available
- No uploading your own voice
- No voice design services
- Eight preset voices only
- No emotion tags or phoneme dictionaries
- No SSML support
Watermarks in Audio
Every audio file generated by our API carries an inaudible watermark that marks it as synthetic speech. This technical marker is embedded directly into the audio data, ensuring that animated content produced via our text-to-speech API can be identified as AI-generated without affecting playback quality.
For animation agencies, this watermark provides a clear distinction between original voice acting and synthesized dialogue. It helps maintain transparency in post-production workflows, especially when licensing assets for broadcast or digital platforms where synthetic origin may need to be disclosed.
Marking Synthetic Speech
The watermark is embedded during the synthesis process and is not visible or audible to the end viewer. It serves as a reliable indicator that the audio was generated from text using our API, rather than recorded by a human voice actor.
This feature is consistent across all eight preset voices and 23 supported languages. It does not require additional processing or metadata tagging after generation, making it seamless for teams managing large volumes of animated dialogue.
- Inaudible to listeners
- Embedded in every generated file
- Works across all voices and languages
- No extra steps needed after synthesis
Cloning vs. Presets for Cartoons
Custom voice cloning captures a specific person’s unique timbre, offering a one-of-a-kind sound. However, our eight preset voices are built on licensed recordings from LibriTTS-R and LibriVox, ensuring consistent, licensable speech without the legal ambiguity of cloning a living actor’s identity.
For animation agencies, presets provide reliable character voices across 23 languages. Unlike cloning, which requires extensive training data and ongoing rights management, our presets are ready to use immediately, allowing you to draft scenes without negotiating individual voice contracts.
Quality Differences
Cloned voices often sound hyper-realistic but can feel uncanny or overly specific to one person. Our presets like Aria, Nova, or Atlas offer distinct, stylized tones ideal for cartoon characters, balancing realism with artistic flexibility. They avoid the 'uncanny valley' effect while maintaining high audio quality.
- Presets offer consistent tone across different scripts and languages
- Cloning requires significant data to avoid robotic artifacts
- Presets are instantly available without training delays
Legal Landscape
Animation agencies rely on clear contracts to secure the rights to use generated speech in commercial projects. Our API delivers licensable speech from eight preset voices, ensuring that agencies own the output without needing complex third-party rights management for individual voice actors.
Contractual Clarity
When using preset voices, the license is straightforward: you pay for the characters used, and the resulting audio is yours to incorporate into your animation. This eliminates the ambiguity often found when hiring human voice actors, where residual payments or usage limits might apply.
- Clear ownership of generated audio files
- No residual payments for preset voices
- Simple licensing for commercial animation projects
When to Choose Cloning
Voice cloning excels when a character requires a unique, signature sound that no preset can replicate, such as a specific celebrity likeness or a highly distinctive personal voice. However, for the vast majority of animation workflows, this level of customization is unnecessary overhead.
If your production relies on standard archetypes—like the brave hero, the wise mentor, or the comedic sidekick—our eight preset voices cover these roles across 23 languages. You get consistent, licensable speech without the complexity of training a new model.
When Presets Suffice
For draft stages and episodic content, preset voices offer the perfect balance of quality and speed. With 4096 characters per request and generation speeds of roughly 25 characters per second of audio, you can iterate through script changes rapidly without waiting for cloning training.
Use presets when you need multiple characters speaking in different languages or when the voice role is defined by tone (e.g., 'calm and instructional') rather than a specific timbre. This approach keeps your production pipeline lean and cost-effective.
- Standard character archetypes (hero, villain, narrator)
- Multi-language dialogue requiring consistent tone
- Rapid script revisions during pre-production
- Background voices and crowd extras

Data Usage in Cloning
ToonVoce does not offer custom voice cloning, so there is no complex data pipeline for training neural models on your specific talent. We rely on eight fixed presets derived from licensed LibriTTS-R recordings, meaning the underlying voice data is static and pre-licensed.
Since we do not store or process unique voice samples for cloning, your input text is used solely to generate the audio file and is not retained for training. This simplifies data usage, as we only keep request metadata for billing purposes, ensuring your script content remains private without the overhead of long-term model training archives.
Storage of Clone Data
Because ToonVoce provides preset voices rather than custom clones, you do not need to worry about where your voice actor’s unique vocal characteristics are stored. Our system uses eight established voice profiles that remain constant, eliminating the need for large-scale storage of individualized voice models or training datasets specific to your productions.
Input text is processed in real-time to produce the audio file. Once the synthesis is complete and the audio is returned, the text is not stored for future training or model improvement. Only minimal request metadata—such as timestamp, character count, and status—is retained for accurate billing, keeping your creative data lean and secure.
- Input text is not used for training
- No custom voice models are stored
- Only billing metadata is retained
- Voice profiles are static and licensed
Transparency in AI Audio
Synthetic speech is becoming ubiquitous in animation, making it crucial to distinguish between human and AI-generated audio. Without clear markers, audiences may struggle to identify the origin of a voice, potentially leading to confusion or unintended attribution of performance to a specific actor.
Our API embeds an inaudible watermark in every generated audio file. This technical signature ensures that producers can verify the synthetic nature of the voice, maintaining transparency with clients and distributors regarding the use of AI-generated dialogue in their projects.
Watermark Importance
The inaudible watermark serves as a digital fingerprint, marking the audio as synthetic without affecting the listening experience. This allows animation agencies to confidently license and distribute content, knowing the origin of every voice track is verifiable.
Unlike human recordings, which rely on metadata or contracts for provenance, watermarked AI audio provides an immediate, technical verification method. This is particularly useful for large-scale productions where thousands of lines of dialogue are generated, ensuring consistent documentation of synthetic assets.
- Inaudible to human ears
- Embedded in every generated file
- Verifies synthetic origin instantly
- Supports clear licensing and distribution
Our Voice Sources
Our eight preset voices are derived from licensed recordings, ensuring your animation projects have clear provenance from the start. We use data from LibriTTS-R, a high-quality dataset of public-domain audiobook readings sourced from LibriVox. This foundation allows us to offer consistent, licensable speech without the legal ambiguity of scraping unverified sources.
By relying on established public-domain material, we provide a transparent base for character voices that animators can trust. The voices are not clones of specific living celebrities but are distinct presets trained on this licensed corpus, ready for immediate use in your production pipeline.
Licensing Verification
Every voice in our API is built on LibriTTS-R data, which is released under the CC BY 4.0 license. This means the underlying recordings are freely available for commercial use with attribution, giving your agency a solid legal footing. You do not need to worry about hidden royalty clauses or complex clearance processes for standard character dialogue.
We ensure that the synthetic speech generated from these sources carries an inaudible watermark. This marks the audio as synthetic, which helps in managing rights and identifying the source of the audio in future audits. It is a simple layer of transparency that complements our open licensing model.
- Built on LibriTTS-R (CC BY 4.0) data
- Sourced from LibriVox public-domain audiobooks
- Clear commercial usage rights for presets
- Inaudible watermark marks synthetic origin
Future of Cloning in Animation
The animation industry is gradually shifting from labor-intensive voice actor sessions toward standardized digital voices. This transition prioritizes consistency across long-running series and rapid production cycles over unique, one-off character nuances. Agencies are finding that preset voices meet the needs of most draft and production stages efficiently.
As technology matures, the gap between cloned and preset audio narrows. However, the cost and complexity of cloning remain barriers for many mid-sized studios. The trend favors scalable, licensable assets that can be deployed across multiple languages without additional recording sessions.
Adoption Rates
Adoption of voice cloning in animation is growing steadily but remains niche compared to preset voice usage. Most agencies prefer presets for their reliability and cost-effectiveness. Cloning is typically reserved for flagship characters or high-budget projects where a specific, recognizable voice is critical.
Current adoption rates indicate that while many studios are exploring cloning, the majority still rely on preset voices. This approach ensures consistent quality across 23 languages without the legal and technical overhead of managing individual voice licenses. The trend suggests a future where presets dominate routine production, with cloning used selectively.
- Most agencies use preset voices for efficiency.
- Cloning is reserved for high-budget projects.
- Preset voices offer consistent quality across languages.
Summary for Agencies
ToonVoce’s eight preset voices deliver consistent, licensable speech across 23 languages, ideal for agencies producing multiple animation drafts. By using established voices like Aria or Felix, teams avoid the cost and delay of hiring voice actors for every prototype.
Our API accepts text via POST requests, returning audio files in formats like MP3 or WAV. With a rate limit of 60 requests per minute, agencies can iterate quickly during the pre-production phase without waiting for human scheduling.
Best Fit for Drafts
Voice cloning requires extensive data collection, consent verification, and legal negotiation, making it costly for temporary character iterations. In contrast, our preset voices provide immediate, high-quality results for storyboarding and animatics.
Agencies benefit from using presets for rough cuts and early reviews. Reserve cloning for final productions where specific brand alignment justifies the additional legal and technical overhead.
- Use presets for rapid prototyping and internal reviews
- Reserve cloning for final commercial releases
- Avoid cloning costs for temporary character variations
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Each voice has a distinct tone suitable for standard adult roles in animation.
23 languages
Any voice speaks any of these languages.
