Is Synthetic Speech the Same as Voice Cloning?

By SpotLogic editorial team · Updated 2026-11-19

This page clarifies the distinction between synthetic speech and voice cloning for advertising agencies producing radio spots. We explain how fixed presets like Aria or Atlas differ from custom cloning, ensuring you understand the licensing and technical realities before integrating our API into your workflow.

Readers will find concrete details on voice licensing, the CC BY 4.0 framework, and the inaudible watermarking used in every generated file. This guide helps you assess whether our eight preset voices meet broadcast standards without the need for custom voice design or additional royalty payments.

Is Synthetic Speech the Same as Voice Cloning?

Frequently asked questions

Is voice cloning the same as voice synthesis?

No. Voice synthesis generates audio from text using preset models, while cloning requires training on specific voice data. We do not offer custom voice cloning; you select from eight fixed presets. Each preset is based on licensed recordings from LibriTTS-R and LibriVox, ensuring clear licensing without the need for individual cloning processes.

Are the voices real people?

The voices are based on licensed recordings from LibriTTS-R, sourced from LibriVox public-domain audiobook readings. While the audio originates from human speakers, the output is synthetic speech. Every generated file carries an inaudible watermark marking it as synthetic, distinguishing it from direct recordings of the original voice actors.

Do I need to pay royalties for the voices?

No additional royalties are required. The voices are based on LibriVox public-domain readings and LibriTTS-R data. You pay only for the API usage per character. The license covers the synthetic generation, allowing you to use the audio in projects without ongoing royalty payments or complex rights management.

Can I clone my own voice?

No. We do not offer custom voice cloning or the ability to upload your own voice. You must choose from the eight available presets: aria, nova, luna, sage, felix, milo, atlas, or orion. Each preset is a fixed model trained on specific source data, ensuring consistent quality and licensing.

How are the voices licensed?

Voices are based on LibriTTS-R (CC BY 4.0) and LibriVox public-domain readings. This provides clear licensing for commercial use. The synthetic audio generated via our API inherits these rights, allowing you to use the output in advertising and other projects without needing separate voice actor agreements or royalty checks.

What is the CC BY 4.0 license?

CC BY 4.0 is a Creative Commons license allowing reuse and adaptation with attribution. Our voices derive from LibriTTS-R, which uses this license. This ensures legal clarity for commercial use. The license permits commercial use, modification, and distribution, provided appropriate credit is given to the original source.

Do the voices sound robotic?

The voices are designed for broadcast quality, offering clear, natural-sounding speech suitable for radio spots. They are based on real human recordings. While synthetic, they avoid the typical robotic tone of older TTS. The inaudible watermark confirms they are AI-generated, but the audio quality is intended for professional advertising use.

Can I use the voices for commercials?

Yes. The voices are licensed for commercial use, making them ideal for advertising agencies producing radio spots and audio ads. You can use any of the eight presets for commercials without additional royalty fees. The synthetic nature is clear, but the quality is suitable for broadcast-ready audio content.

Are there any restrictions on usage?

Usage is primarily restricted by rate limits (60 requests per minute) and character limits (4096 per request). There are no content restrictions on the text itself. You cannot clone your own voice. Input text is not used for training. The audio is synthetic and watermarked.

How long does the audio generation take?

Generation is roughly real-time, about 25 characters per second. Short phrases take a few seconds. Longer texts are split and joined. Requests are processed sequentially, so under load, you may experience queue delays. This timing applies to both tts-1 and tts-1-hd models, with no significant speed difference.

Can I change the voice after generation?

No. Once the audio file is generated, the voice is fixed. You cannot alter the voice preset in the resulting file. To change the voice, you must send a new request with the desired voice parameter. The API does not support post-generation voice modification or editing of the audio structure.

Do I own the generated audio?

You own the generated audio files you create. The input text is not used for training. We do not store the generated audio. You retain rights to use the audio in your projects. Generated audio is not stored by the system.

Is the audio watermarked?

Yes. Every generated file carries an inaudible watermark that marks it as synthetic speech. This watermark is embedded in the audio data and is not audible to listeners. It serves to identify the audio as AI-generated, which is useful for transparency and content verification in broadcast and digital media.

What is the difference between tts-1 and tts-1-hd?

Both models use the same engine and price. The difference is audio quality: tts-1-hd offers higher fidelity. Both accept the same parameters, voices, and formats. There is no speed difference, and both are processed at the same rate. Choose tts-1-hd for higher fidelity, or tts-1 for standard needs.

Definition of Synthetic Speech

Sound wave illustration for: Definition of Synthetic Speech

Synthetic speech, or text-to-speech (TTS), converts written text into audio files using cloud API endpoints. Unlike human voice actors, these voices are not recordings of specific individuals but algorithms generating sound from data. The process involves sending text to a server, which returns synthesized audio without human intervention.

Text-to-Speech Technology

TTS technology reads input text and synthesizes it into audible speech. The system processes characters through a neural network, outputting audio files in formats like MP3 or WAV. This method allows agencies to generate radio spots instantly without booking studio time or actors.

AI-Generated Speech

AI-generated speech uses machine learning models trained on large datasets. The models learn pronunciation and intonation patterns to produce natural-sounding audio. Each request is processed individually, ensuring that the output is unique to the input text provided.

Cloning vs. Fixed Presets

Our service uses fixed presets rather than custom voice cloning. You select from eight licensed voices, such as Aria or Atlas, which are based on specific recordings. Cloning involves training a model on a unique voice sample; here, the voices are standardized assets available to all users.

How It Works

Sound wave illustration for: How It Works

SpotLogic’s text-to-speech API converts your script into audio through a streamlined pipeline. The process begins by selecting from eight fixed voice presets, which serve as the foundational timbre for the output.

The engine then processes the text input, applying speed and format parameters before generating the final audio file. This ensures consistent, broadcast-ready results without the variability of live recording sessions.

Each synthesis request is handled sequentially, with typical generation speeds mirroring real-time speech. The system supports multiple audio formats (mp3, wav, opus, flac, pcm), allowing agencies to choose the optimal file type for their specific radio or digital ad workflows.

Voice Presets

The API uses eight distinct voice presets, including female options like Aria and Nova, and male options like Felix and Atlas. Each preset is based on licensed recordings from the LibriTTS-R dataset, ensuring clear, consistent tonal qualities.

These presets are not cloned from specific individuals but are engineered profiles that can speak any of the 23 supported languages. This allows agencies to maintain brand voice consistency across different campaigns without negotiating individual talent contracts.

  • Female: aria, nova, luna, sage
  • Male: felix, milo, atlas, orion
  • Each voice supports all 23 languages

Engine Processing

When you send a request, the engine validates the input text against the 4096-character limit and applies the specified speed and language parameters. The system processes requests sequentially, which may introduce slight delays during high-load periods.

Authentication is handled via a single API key, ensuring secure access to the synthesis engine. The engine does not store the generated audio, returning it directly in the response to minimize data retention.

Rate limits are set at 60 requests per minute, with a lower threshold of 10 per minute for new accounts before their first top-up. This structure encourages efficient batching of scripts for production workflows.

  • Sequential processing ensures stability
  • No audio storage after generation
  • 60 requests per minute limit

Audio Synthesis

The synthesis engine generates audio files in formats such as mp3, wav, opus, flac, and pcm. Each file includes an inaudible watermark to identify it as synthetic speech, which helps in tracking usage and verifying authenticity.

The output is delivered as a direct audio file response, ready for immediate integration into radio spots or digital ads. The generation speed is approximately real-time, meaning a one-minute audio file takes about one minute to synthesize.

Failed requests are free, so agencies can test different voice settings or text lengths without financial risk. This allows for rapid iteration during the creative process, ensuring the final audio meets the required broadcast standards.

  • Formats: mp3, wav, opus, flac, pcm
  • Inaudible watermark included
  • Failed requests are free

Licensing of Voices

Sound wave illustration for: Licensing of Voices

SpotLogic’s voice models are built on the LibriTTS-R dataset, which provides the foundational audio samples for our eight preset voices. This approach ensures that every synthetic voice is rooted in verified, high-quality recordings rather than abstract neural patterns.

The licensing framework is transparent and standardized, allowing advertising agencies to integrate these voices into radio spots and audio ads without ambiguity regarding ownership or usage rights.

LibriTTS-R Dataset

The core of our voice engine is the LibriTTS-R dataset, a curated collection of speech recordings. These recordings serve as the direct source material for the eight voice presets, from Aria to Orion, ensuring consistent and broadcast-ready audio quality.

Each preset is derived from specific recordings within this dataset, maintaining a clear lineage from the original performance to the final synthetic output used in your productions.

CC BY 4.0 License

The voices are based on recordings licensed under CC BY 4.0. This license permits commercial use, modification, and distribution, provided appropriate credit is given to the original source. This simplifies legal review for agencies producing paid advertising.

You can confidently use these voices in commercial campaigns without worrying about complex royalty structures or restrictive exclusive licensing terms that often plague traditional voice actor contracts.

Public Domain Sources

The underlying recordings originate from LibriVox, a project that aggregates public domain audiobook readings. The recordings originate from LibriVox public-domain audiobook readings and are available for reuse under the CC BY 4.0 terms.

This public domain foundation means the voices are not tied to exclusive performer contracts, offering agencies a stable, predictable cost structure for ongoing advertising production needs.

Consent and Rights

Sound wave illustration for: Consent and Rights

Our synthetic voices are generated from the LibriTTS-R dataset, comprising recordings from LibriVox public-domain audiobook readings. These sources are primarily public-domain audiobook readings, ensuring a clear chain of rights for advertising agencies.

The CC BY 4.0 license governing the source material allows for commercial use without additional royalty payments. This means your generated audio files are ready for broadcast without ongoing licensing fees, simplifying rights management for your campaigns.

Source Material Consent

The voices originate from the LibriTTS-R dataset, derived from LibriVox public-domain audiobook readings. This foundational consent ensures that the synthetic outputs do not infringe on personal rights of identifiable individuals.

Commercial Use Rights

Covered by the CC BY 4.0 license, the source material permits commercial usage. Agencies can generate radio spots and audio ads without paying additional royalties or negotiating separate rights for each voice preset used.

  • No per-use royalties required
  • Commercial use permitted under CC BY 4.0
  • Clear rights chain from volunteer to final audio

Watermarking

Every generated audio file contains an inaudible watermark marking it as synthetic speech. This provides transparency and helps identify AI-generated content in broadcast or digital environments without affecting the listening experience.

Use Cases in Advertising

Sound wave illustration for: Use Cases in Advertising

SpotLogic enables agencies to produce broadcast-ready audio without scheduling studio time. By sending text to the API, teams generate high-quality synthetic voices for radio spots and audio ads instantly.

The API supports 23 languages, allowing global campaigns to localize content rapidly. With predictable pricing and no minimum spend, agencies can scale production from a single spot to a full campaign.

Radio Spots

Generate voiceovers for radio commercials in seconds. The engine handles up to 4,096 characters per request, covering standard ad lengths. Voices like Aria or Felix provide clear, engaging narration suitable for broadcast.

  • Instant generation for rapid campaign iterations
  • Support for 23 languages including English, Spanish, and German
  • Consistent tone across multiple ad variants

IVR Systems

Create natural-sounding interactive voice response menus. Use calm voices like Sage or Milo for instructional guidance. The API returns audio files in MP3, WAV, or Opus formats, ready for integration into existing telephony systems.

  • Dedicated voices for instructional and conversational tones
  • Multiple audio formats for flexible integration
  • No streaming required; static audio files

E-Learning

Produce educational content with clear, neutral narration. Atlas or Luna provide steady, readable delivery for training modules. Since input text is not used for training, your proprietary course material remains secure.

  • Neutral voices for clear instruction
  • Secure input text handling
  • Scalable production for large courses

Benefits for Agencies

Sound wave illustration for: Benefits for Agencies

SpotLogic eliminates the recurring costs of booking voice talent for every spot. With a flat rate of $0.01 per 1,000 characters, agencies can scale production without negotiating per-project fees or managing union rates for minor tweaks.

Cost Efficiency

Failed synthesis requests are free, ensuring you only pay for usable audio. The predictable per-character pricing allows precise budget forecasting for high-volume campaigns, replacing unpredictable hourly studio rates with a fixed, low-cost model.

  • Flat rate: $0.01 per 1,000 characters
  • Zero cost for failed requests
  • No minimum spend or monthly subscriptions

Speed of Production

Generate audio files in seconds rather than scheduling studio time. Longer texts process at roughly real-time speed, while short phrases take only a few seconds. This instant turnaround accelerates revision cycles and shortens campaign launch windows significantly.

  • Instant generation for short phrases
  • Real-time speed for longer texts
  • No studio scheduling delays

Consistency

Every voice preset delivers uniform tone across all your content. Whether producing a 10-second radio spot or a 5-minute e-learning module, the vocal character remains identical, maintaining brand integrity without the variability of different human actors.

  • Uniform vocal tone across all assets
  • No variability between different actors
  • Brand consistency maintained automatically

Limitations

Sound wave illustration for: Limitations

Synthetic speech offers speed and consistency, but it lacks the flexibility of human performers. Agencies must accept the constraints of the current engine without the ability to tailor voices to specific brand needs.

These limitations are structural to the API design. Understanding them helps set realistic expectations for audio ad production cycles.

No Custom Cloning

The service does not offer voice cloning. You cannot upload a reference recording to create a unique voice profile for your brand.

This means every spot uses one of the eight available presets. There is no option to generate a bespoke voice from scratch.

Fixed Presets

All audio is generated using eight fixed voice presets. You cannot adjust individual parameters like pitch, timbre, or emotion tags.

This limits creative variation but ensures consistent, predictable results across high-volume production runs.

Watermarked Output

Every generated audio file contains an inaudible watermark identifying it as synthetic speech.

This marks the content as AI-generated but does not affect playback quality or broadcast readiness.

Limitations

Voice Selection

Sound wave illustration for: Voice Selection

Choosing the right voice is critical for brand recognition in advertising. SpotLogic provides eight distinct presets, each engineered for specific tonal qualities that resonate with different audience segments.

Agency directors can select from four male and four female voices, ensuring precise alignment with the campaign’s narrative voice without the variability of human actors.

Male Presets

Felix delivers energetic, promo-style delivery ideal for high-impact radio spots. Milo offers a friendly, conversational tone suitable for direct-to-consumer messaging.

Atlas provides a neutral, news-style voice for clear information delivery. Orion’s deep, documentary tone adds gravitas to premium brand stories.

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

Female Presets

Aria serves as a warm, clear narrator, perfect for storytelling. Nova brings a bright, upbeat energy to lifestyle advertisements.

Luna offers a soft, storytelling voice for emotional connections. Sage provides a calm, instructional tone for explanatory content.

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

Tone Matching

Each voice is designed for specific advertising contexts. Select based on the emotional arc of your script rather than just gender.

Use the /v1/voices endpoint to preview tone characteristics. Match the preset to your brand’s existing audio identity for consistency.

Technical Requirements

Sound wave illustration for: Technical Requirements

Integration requires a single POST request to the API endpoint, utilizing standard JSON bodies compatible with existing OpenAI SDKs. Developers simply adjust the base URL to https://api.ah777.info/v1 and provide their Bearer token for authentication, ensuring seamless adoption into current advertising production pipelines without custom middleware.

API Integration

Authentication is handled via a Bearer token in the Authorization header. The endpoint accepts standard JSON payloads, allowing agencies to use familiar libraries like the official OpenAI SDK by merely updating the base URL configuration.

  • Base URL: https://api.ah777.info/v1
  • Method: POST
  • Auth: Bearer token

Text Input

Send your script as the input field within the JSON body, capped at 4096 characters per request. Longer scripts require splitting into segments and joining the resulting audio files. The language field is optional but recommended for non-English Latin scripts to ensure accurate pronunciation.

  • Max 4096 characters per request
  • Supports standard text encoding

Audio Output

The API returns the synthesized audio file directly in the response body. You can specify the format using the response_format parameter, choosing from mp3, wav, opus, flac, or pcm to match your broadcast or digital distribution standards.

  • Formats: mp3, wav, opus, flac, pcm
  • Direct file return in response

Comparison with Human Voice Actors

Sound wave illustration for: Comparison with Human Voice Actors

Synthetic speech offers a distinct alternative to human voice actors for advertising agencies, primarily through lower costs and faster turnaround. While human actors require booking, studio time, and editing, the API delivers broadcast-ready audio instantly via simple API calls.

Cost Comparison

Human voice talent involves session fees, union rates, and post-production costs that vary by project scope. SpotLogic charges $0.01 per 1,000 characters, providing a predictable, usage-based model without minimums or subscription fees. This eliminates the overhead of studio rentals and editing hours.

  • No session fees or union rates
  • Pay only for characters generated
  • No minimum spend requirements

Speed Comparison

Hiring an actor involves casting, scheduling, and recording sessions that can take days. Synthetic speech generates audio in seconds, with typical processing speeds matching real-time. This allows agencies to iterate on scripts rapidly without waiting for talent availability or studio slots.

  • Instant audio generation
  • No scheduling delays
  • Rapid script iteration

Quality Comparison

Human voices offer nuanced emotional expression, while SpotLogic provides eight distinct presets optimized for clarity and broadcast standards. The synthetic voices are based on licensed recordings, ensuring consistent quality across projects. Each file includes an inaudible watermark to identify synthetic content, which human recordings lack.

  • Consistent broadcast-ready quality
  • Eight distinct voice presets
  • Inaudible synthetic watermark

Future of Synthetic Speech

Sound wave illustration for: Future of Synthetic Speech

Synthetic speech is steadily replacing traditional voiceovers for routine advertising assets. Agencies are integrating TTS into their production pipelines to handle high-volume, low-complexity spots efficiently.

Industry Trends

Brands are prioritizing speed over bespoke nuance for digital ads. The shift toward API-driven audio allows agencies to scale content without negotiating new talent contracts for every variation.

Adoption Rates

Adoption is growing among mid-sized agencies seeking predictable costs. The model’s fixed pricing encourages testing multiple voices without the overhead of studio bookings.

Technological Advances

Recent improvements in voice presets like Aria and Atlas offer clearer, more natural narration styles. While no new features are promised, the current engine delivers broadcast-ready quality for standard spots.

FAQs on Synthetic Speech

Sound wave illustration for: FAQs on Synthetic Speech

Synthetic speech generates audio from text using fixed presets, while voice cloning recreates a specific individual’s unique vocal characteristics. Our system uses licensed recordings of professional voice actors, ensuring clear licensing and predictable pricing without the need for custom cloning.

Each voice preset is trained on the LibriTTS-R dataset, providing consistent, broadcast-ready quality. This approach eliminates the variability of human actors while maintaining high production standards for advertising agencies.

Is voice cloning the same as voice synthesis?

No. Voice synthesis converts text to speech using predefined voice models. Voice cloning creates a unique voice model based on specific recordings of an individual. Our API uses fixed presets, not custom cloning.

Synthetic speech relies on trained voice presets that can speak any supported language. Cloning requires specific audio samples of a target person to replicate their exact vocal nuances.

Are the voices real people?

Yes, the voices are based on licensed recordings from the LibriTTS-R dataset. These recordings originate from LibriVox public-domain audiobook readings, ensuring authentic human vocal qualities.

Each voice preset, such as Aria or Felix, is derived from real voice actors. The synthesis engine processes these recordings to generate natural-sounding speech from text input.

Do I need to pay royalties for the voices?

No. The voices are licensed under CC BY 4.0, allowing commercial use without ongoing royalties. Once you generate audio, you own the output without additional licensing fees.

The CC BY 4.0 license permits reuse, modification, and commercial application. You only pay for the character processing, not for recurring voice usage fees.

Can I clone my own voice?

No. We do not offer custom voice cloning or uploading of personal voice samples. You must choose from the eight available voice presets.

Our system uses fixed models trained on existing datasets. While this ensures consistency, it does not allow for creating unique voice profiles from custom recordings.

How are the voices licensed?

The underlying voice data is licensed under CC BY 4.0 from LibriVox. This allows commercial use with proper attribution if required by the source.

You can use the generated audio for any purpose, including radio spots and ads, without additional licensing constraints. The license covers the voice model, not the specific generated file.

What is the CC BY 4.0 license?

CC BY 4.0 is a Creative Commons license that allows free use, adaptation, and commercial application of the voice data. It requires attribution to the original source.

This license ensures that the voice presets are legally usable for commercial advertising. It provides clarity on rights, avoiding complex royalty structures common in traditional voice acting.

Do the voices sound robotic?

No. The voices are based on real human recordings, providing natural intonation and emotion. They are designed for broadcast-ready quality in advertising.

While synthetic, the voices avoid the robotic tone of older TTS systems. They offer clear, professional delivery suitable for radio spots and promotional content.

Can I use the voices for commercials?

Yes. The voices are ideal for commercials, radio spots, and IVR systems. They provide consistent, high-quality audio for advertising campaigns.

Each voice preset offers distinct characteristics, allowing you to match the tone to your brand. The output is ready for immediate use in commercial productions.

Are there any restrictions on usage?

The main restriction is the character limit of 4096 per request. Longer texts must be split and joined. There are no content restrictions beyond standard copyright.

You can use the audio for any commercial or non-commercial purpose. The generated files are not stored, ensuring privacy and flexibility in usage.

How long does the audio generation take?

Generation takes roughly real-time, about 25 characters per second. Short phrases may take a few seconds, while longer texts process proportionally.

Requests are processed sequentially, so under load, there may be slight delays. The system ensures high quality without compromising on speed for typical advertising needs.

Can I change the voice after generation?

No. Once audio is generated, the voice is fixed. You can regenerate the audio with a different voice preset if needed, but the original file remains unchanged.

Each request produces a unique audio file based on the selected voice. Changing the voice requires a new API call with the desired voice parameter.

Do I own the generated audio?

Yes. You own the generated audio files. The input text is used only for synthesis and not stored, ensuring your content remains yours.

The generated audio is returned in the response and not stored by our system. You have full rights to use the audio in your advertising campaigns.

Is the audio watermarked?

Yes. Each generated file carries an inaudible watermark to indicate it is synthetic speech. This helps identify AI-generated content without affecting the audio quality.

The watermark is inaudible to listeners, ensuring a professional presentation. It serves as a marker of synthetic origin, useful for transparency in media production.

What is the difference between tts-1 and tts-1-hd?

Both models use the same engine and voice presets. The difference lies in the audio quality, with tts-1-hd offering higher fidelity.

Pricing is identical for both models. Choose tts-1-hd for superior audio clarity, especially for high-end radio spots or premium advertising content.

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. Selecting the right format depends on your final distribution channel.

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. Tether (USDT)
  3. OpenAI API reference: create speech
  4. curl documentation
  5. Creative Commons Attribution 4.0
  6. LibriTTS-R corpus (OpenSLR 141)
  7. MDN: Web Speech API
  8. LibriVox public-domain audiobooks
  9. TRON developer hub: TRC-20
  10. W3C WAI: making audio and video accessible
  11. FFmpeg documentation
  12. Node.js documentation

More on this site