SYNTX AI

RU|EN

🎙️ ElevenLabs Voice

9 min read

ElevenLabs Voice is a tool for generating realistic speech from text or audio, delivering natural sound and highly accurate intonation. It works well for both everyday tasks and professional use.

🎛 Navigation:

🏠 Main Menu > 🔊 AI Audio > 🎙️ ElevenLabs Voice

💥 Step-by-step guide to using the “Text to Speech” mode

This mode converts written text in over 70 languages into natural-sounding speech using more than 200 different voices. You can choose from a wide selection of voices with different accents, suitable for various tasks, all designed to sound as natural and lifelike as possible.

❣️ Maximum number of characters per request - 4090.

❣️ In rare cases, the AI may change the tone of voice in the middle of the audio. 

Unfortunately, this cannot be controlled. If this happens, we recommend simply regenerating the audio.

First, send the text you want to convert into speech. Then click the “Choose Voice” button to select a voice for the generation.

❣️ The system will show the number of characters and the final cost of the generation in advance!

Example request:

After selecting a voice, you will have the option to delete the request, choose another voice, or start the synthesis.

Delete request
This button completely clears the entered text, allowing you to start over.

Choose another voice
This button opens the voice library, where you can change the speaker, choose a different accent, or select another voice style for your audio.

Start synthesis
This button starts the audio generation process based on the entered text and selected settings.

Result:

💥 Step-by-step guide to using the “Speech to Speech” mode

This mode allows you to transform the original voice into another one while preserving its tone and delivery style. You can use either an uploaded audio file or a microphone recording as input.

❣️ Maximum uploaded audio file size - 20 MB.

❣️ In rare cases, the AI may change the tone of voice in the middle of the audio. Unfortunately, this cannot be controlled. If this happens, we recommend simply regenerating the audio.

Key features of the “Speech to Speech” mode:

• High accuracy in reproducing whispers and quiet speech
• Ability to include natural sounds such as breaths, laughter, or crying
• Accurate recognition and transfer of tone and emotions
• Precise preservation of speech rhythm and cadence
• Retention of the original language and accent

First, upload the audio you want to work with. Then click the “Choose Voice” button to choose a voice for the upcoming generation.

Example request:

After selecting a voice, you will have the option to delete the request, choose an another voice, or start the synthesis.

Delete request
This button completely clears the entered text, allowing you to start over.

Choose another voice
This button opens the voice library, where you can change the speaker, choose a different accent, or select another voice style for your audio.

Start synthesis
This button starts the audio generation process based on the entered text and selected settings.

Original version:

Result:

💥 Step-by-step guide to using the “Text-To-Dialogue” mode

If you need to generate a voiced dialogue, you can switch to the “Text-To-Dialogue” mode. In this mode, the tool processes dialogue structure and punctuation more accurately, resulting in more natural-sounding speech.

First, in the voice library, select the “Text-To-Dialogue” model.

After that, send your dialogue to the chatbot and select the language for the voice generation.

Example request:

After selecting a voice, you will have the option to delete the request, choose another voice, or start the synthesis.

Result:

Voice Base

The voice base is divided into two groups: Premium and PRO. All sections are available starting from the BASIC subscription.

Premium voices

In this section, you can filter voices by age, gender, and application to find the most suitable option for your task.

Gender:

  • Male
  • Female

Age:

  • Middle-aged
  • Young

Application:

  • For advertising
  • Voiceover for characters
  • Conversational voices
  • For entertainment
  • For education
  • For narratives
  • For social media

PRO voices

In this section, you can filter voices by name, age, gender, application and accent to precisely find the most suitable option.

Gender:

  • Male
  • Female
  • Neutral

Age:

  • Young
  • Middle-aged
  • Elderly

Application:

  • For advertising
  • Voiceover for characters
  • Conversational voices
  • For entertainment
  • For education
  • For narratives
  • For social media

Accent:

You can choose from 118 different accents. They let you adjust pronunciation and voice characteristics, making the speech sound more authentic and better suited to a specific region or language environment.

              african, african american, american, andalusian, arabic, argentine, athenian, australian, bavarian, beijing mandarin, bengali, berlinerisch, bihari, boston, brabantian, brazilian, british, bucovina, budapest, calabrese, canadian, canary islands, cebuano, central, chennai, chicago, chilean, chinese, colombian, creole, croatian, cuban, cypriot, danish, dutch, egyptian, european, filipino, flemish, french, galician, german, gothenburg, greek, gujarati, gulf, gyeongsang, haryanvi, helsinki, hindi, hong kong cantonese, indian, irish, istanbul, italian, jamaican, japanese, javanese, jeolla, kansai, kanto, kyushu, latin american, levantine, madeiran, malay, marathi, mazovian, mexican, milanese, modern standard, moscow, new york, new zealand, nigerian, northern, oslo, parisian, peninsular, peruvian, polish, portuguese, prague, puerto rican, punjabi, quebec, received pronunciation, rhine franconian, romanesco, romanian, russian, saint petersburg, saudi, scottish, scouse, seoul, sicilian, singaporean, south african, southern, spanish, standard, stockholm, swabian, swedish, taiwan mandarin, tamil, tunisian, turkish, tuscan, ukrainian, us midwest, us southern, venezuelan, welsh, western, yorkshire, zagreb

Search by name:

This filter allows you to quickly find a specific voice by its name if you remember it.

🔀 Sort by

Sort by rating:

Displays the highest-rated voices, helping you quickly find the best options and save time.

Latest additions:

Shows recently added or updated voices, so you can easily keep track of new arrivals.

From A to Z:

Ideal when you know the exact name. Provides quick and easy access to the desired voice.

From Z to A:

Reverses the list order, allowing you to browse from the end and discover less obvious options.

❤️ Favorites

If you find a voice you like, you can add it to the “Favorites” section by clicking the “like” icon next to the selected voice.
Saved voices are stored in a separate section, allowing you to quickly find and use them later.

⚙️ Options

Options allow you to adjust the generation settings: control playback speed, voice stability, similarity to the original voice, and add stylistic exaggeration for more expressive speech.

Audio speed

Controls the speech playback rate — how fast or slow the voice delivers the text. This parameter can be adjusted to suit your needs, making the speech more dynamic or, conversely, calmer and clearer.

Stability

Controls how smooth and consistent the voice sounds. Lower values make the speech more expressive and emotional, with greater variation in pitch and intonation. Higher values make the voice more even and stable, but it may also sound more monotone. Very low values can cause abrupt and unpredictable changes.

Similarity

Determines how closely the generated voice matches the original. Higher values preserve more of the original voice’s characteristics, including possible recording artifacts. Lower values produce a cleaner sound, but some of the original voice’s unique traits may be lost.

Style exaggeration

Enhances the distinctive characteristics and expressiveness of the original speaker’s voice. Increasing this value requires more resources and may slow down generation. In most cases, it is recommended to keep this parameter at 0.

  • Stability is most often set around 50, and similarity around 80. These values usually provide a balanced result, after which further adjustments tend to be minor.

  • For a more lively and dramatic sound, it is recommended to lower the stability and generate several variations to choose the best one.

  • If a more controlled and even tone is needed, closer to monotone, it is recommended to increase the stability value.

💡 General recommendations

  • When using numbers in digit form, playback quality may decrease, so it is recommended to write them out in words.

  • When composing text, it is important to follow proper spelling and punctuation rules. This directly affects the quality of the generated speech.

❌ Incorrect:

a christmas tree decorated with sparkling garlands stood right in the center of the living room

✅ Correct:

A Christmas tree, decorated with sparkling garlands, stood right in the center of the living room.

💡 Tips for pauses and emphasis

To indicate emphasis, it is recommended to elongate the desired letter using a triple repetition. Since the system does not have a built-in stress marking feature, the result may not always be perfect.

Example:

In the silence of the niiiight, on an empty street, a loud sound of breaking glass suddenly echoed.

To indicate pauses in the text, it is recommended to use ellipses (…). It is important to follow proper spelling and punctuation rules, as this directly affects the quality of the audio.

Example:

In a dimly lit room, a woman sat… Suddenly, in complete silence, there was a sound - a quiet, barely noticeable creak of the floor behind her.

💡 Use cases:

  • Education and learning: the tool can be used to create interactive learning materials, audiobooks, and educational courses.

  • Video games and entertainment: the tool allows you to create unique voices and dialogues, making content more immersive and engaging.

  • Marketing and advertising: the tool helps generate personalized audio messages to increase audience engagement.

  • Audiobooks: the tool makes it easy to convert text into audio, making literature accessible to a wider audience, including people with special needs.

  • Podcasts: the tool simplifies content creation, especially if recording your own voice is difficult or multiple speakers are required.

  • Accessibility: the tool helps make information more accessible for people with visual impairments and other limitations.

Thank you for taking the time to review this guide. We are confident that the knowledge you’ve gained will help you use the tool effectively and achieve great results in your generations. We wish you the best of luck and plenty of inspiration in your future creative work! 💛

SYNTX AI: Syntx AI
SYNTX Support: Syntx Support YouTube channel “SYNTX”: Syntx YouTube SYNTX.AI Academy: SYNTX Academy

Graph View