Frequently asked questions
What is the difference between tts-1 and tts-1-hd?
Both models use the same engine and price. You can choose either 'tts-1' or 'tts-1-hd' in your request. The audio quality and generation speed are identical, so you can pick your preferred label without affecting cost or performance.
Which voice is best for historical tours?
For historical tours, Orion is suitable for its deep, documentary style, while Atlas provides a neutral, news-like tone. Sage offers a calm, instructional voice suitable for detailed explanations. Choose the preset that best matches the specific atmosphere of your museum exhibits.
Can I change the speed of the voice?
Yes, you can adjust the speed between 0.25 and 4.0. This allows you to slow down complex historical details or speed up shorter exhibits. The API accepts a speed parameter in your request to customize the playback rate for your audience.
Do voices sound different in other languages?
No, every voice preset can speak any of the 23 supported languages. The voice character remains consistent regardless of the language. You simply specify the ISO language code, and the same voice will deliver the content in that language.
Is Aria a female or male voice?
Aria is a female voice described as warm and clear, ideal for narration. It is one of four female presets available. Use it for engaging storytelling or clear instructions within your museum audio guide scripts.
Can I mix multiple voices in one tour?
The API generates one audio file per request. To mix voices, you must generate separate audio files for each voice and join them. There is no built-in feature to blend voices within a single generation request.
What does the language parameter do?
The language parameter sets the spoken language using an ISO code. If omitted, the system guesses the language from the text script, defaulting to English for Latin script. For non-English text, always set this parameter to ensure correct pronunciation.
Are the voices based on real people?
Yes, the voices are based on licensed recordings from LibriTTS-R, derived from LibriVox public-domain audiobook readings. Each generated file also contains an inaudible watermark marking it as synthetic speech, confirming its origin as AI-generated audio.
Can I clone my own voice?
No, custom voice cloning is not offered. You cannot upload your own voice or design a new one. You must choose from the eight available presets: Aria, Nova, Luna, Sage, Felix, Milo, Atlas, or Orion.
Which format is best for mobile devices?
MP3 is widely supported and efficient for mobile devices. You can also use WAV, OPUS, FLAC, or PCM. MP3 is generally preferred for its balance of quality and file size, ensuring smooth playback on most museum guide apps.
How do I specify a German accent?
Set the 'language' parameter to 'de' (German). The voice will then speak German using the selected preset's tone. For example, Atlas will deliver a neutral German news style, while Felix will provide an energetic German promo style.
Is the audio quality better with HD?
No, 'tts-1-hd' uses the same engine as 'tts-1'. There is no difference in audio quality or generation speed. Both models are priced identically at $10 per million characters, so the choice is purely based on your naming preference.
Can I use Felix for a calm tour?
Felix is energetic and promo-oriented, so Sage (calm, instructional) or Luna (soft, storytelling) are better for calm tours. However, you can still use Felix by adjusting the speed parameter (0.25–4.0) to slow it down. This allows you to adapt his upbeat tone to a serene museum environment if preferred.
Do voices have background music?
No, the API generates speech only. There is no background music, sound effects, or SSML support. You must add music separately in your audio editing software or app after receiving the generated audio file from the API.
Understanding voice presets
This page helps tourism and museum staff pick from eight distinct presets without coding. We compare neutral, engaging, energetic, and deep styles so you can match your exhibit’s tone.
Whether you need clear historical facts or warm storytelling, we show which preset fits best.
8 distinct voice styles available
You have eight unique voice presets to choose from. Each is designed for a specific delivery style, ensuring your audio guide sounds professional and appropriate for the content.
The options cover a wide range of tones, from calm instructional guides to energetic promotional clips. You can select the voice that best complements your exhibit’s atmosphere.
Male and female options
The library includes four female and four male voices. This variety allows you to diversify your audio guides or stick to a consistent gender for a single tour.
You can assign different voices to different sections of your museum to keep the experience engaging for visitors.
Each voice has a unique tone
Every voice has a distinct character, such as the warm clarity of Aria or the deep authority of Orion. This specificity helps you match the audio to the subject matter.
Using the right voice enhances listener engagement and makes historical narratives or factual information more digestible.
Neutral voices for facts
When presenting factual historical data, neutral voices provide clarity without distracting from the information. Museum audio guides often require a straightforward delivery that allows visitors to absorb dates, names, and events without emotional interference.
Two presets are suitable for this specific use case, ensuring your content remains the focal point of the experience.
Atlas for news-style delivery
Atlas offers a neutral, news-style delivery ideal for straightforward historical facts. Its balanced tone ensures that dates and locations are communicated clearly, making it suitable for exhibits where accuracy and neutrality are prioritized over narrative flair.
Use Atlas when the content is primarily informational, such as timelines or artifact descriptions, where the voice should remain unobtrusive.
- Ideal for factual historical data
- Neutral, balanced delivery
- Unobtrusive for information-heavy exhibits
Sage for instructional content
Sage provides a calm, instructional tone perfect for guiding visitors through specific processes or exhibits. Its steady pace helps listeners follow along with complex layouts or detailed instructions without feeling rushed.
This preset works well for museum tours that require clear direction, such as navigating large halls or understanding interactive displays.
- Calm, instructional tone
- Steady pace for clarity
- Good for guided navigation
Clear and professional tone
Both Atlas and Sage maintain a professional standard that respects the dignity of museum content. Their consistent delivery ensures that visitors receive information in a respectful and engaging manner, enhancing the overall educational value of the audio guide.
By choosing a neutral voice, you avoid adding unintended emotional layers to factual content, allowing the history to speak for itself.
- Professional and respectful delivery
- Consistent tone across exhibits
- Enhances educational value
Engaging voices for stories
Stories bring history to life, and the right voice makes the difference between a dry recitation and an engaging experience. Aria and Luna offer warmth that invites visitors into the narrative, while Nova adds a bright energy suitable for dynamic historical anecdotes.
Aria for warm narration
Aria delivers a warm, clear narrator style ideal for emotional historical stories. Her tone feels personal and inviting, helping visitors connect with the past on a human level.
- Warm and clear narrative tone
Nova for upbeat energy
Nova brings bright, upbeat energy to your audio guides. She is perfect for lively anecdotes or sections where you want to keep visitors engaged and excited about what they are learning.
- Bright and upbeat delivery
Luna for storytelling style
Luna specializes in a soft, storytelling style that draws listeners in. Her gentle pace is excellent for immersive historical narratives, allowing the story to unfold naturally without rushing the audience.
- Soft, immersive storytelling
Energetic voices for promos
Promotional clips benefit from high-energy audio that captures attention quickly. Felix delivers a dynamic, upbeat tone ideal for short announcements or introductory segments within your museum or tour experience.
This preset is designed to convey enthusiasm without sounding artificial, making it suitable for highlighting special exhibitions or limited-time offers to visitors.
Felix for dynamic content
Felix provides an energetic, promo-style voice that stands out. His clear, bright delivery ensures your promotional messages are heard and remembered.
Use this voice when you want to inject energy into standard guide scripts, transforming routine information into engaging audio clips.
- Ideal for short, punchy announcements
- Best for highlighting special events
- Clear and enthusiastic delivery
High energy delivery
The high-energy tone of Felix keeps listeners engaged during brief segments. It prevents audio fatigue by varying the pace and tone from standard narration.
This style works well when you need to convey excitement or urgency, such as announcing a tour departure or a new exhibit opening.
Good for introductory clips
Introductory clips often set the tone for the entire experience. Felix’s upbeat nature makes these moments memorable and welcoming for new visitors.
Use this preset for welcome messages that introduce the museum or tour theme, ensuring a positive first impression through sound.
Deep voices for documentaries
Orion offers a deep, documentary-style voice suitable for serious historical narratives. His authoritative tone lends weight to complex historical facts without sounding overly dramatic.
This preset is particularly effective for sites requiring a respectful, contemplative atmosphere, ensuring the audio guide feels like a professional broadcast rather than a casual chat.
Orion for serious tone
Orion’s deep register provides a grounded, serious tone suitable for weighty historical subjects. His delivery avoids lightness, making him ideal for solemn moments or significant historical events where a lighter voice might feel out of place.
Deep and authoritative
The authoritative nature of Orion’s voice commands attention, making complex historical details easier to follow. His consistent pacing and deep resonance help maintain listener engagement during longer, information-heavy segments.
Unlike brighter voices, Orion’s timbre conveys stability and trust, which is crucial when presenting verified historical data or museum artifacts.”],
- Consistent deep resonance
Suitable for historical sites
Historical sites often benefit from Orion’s documentary style, which mimics the tone of high-quality audio documentaries. This creates a cohesive experience where the audio guide feels like an extension of the physical environment.
Use Orion for deep dives into history, architectural details, or biographical accounts of significant figures, ensuring the content’s gravity is matched by the voice’s character.”],
Matching voice to content
Aligning the voice preset with your specific content type ensures the audio guide feels natural rather than mechanical. For instance, using a warm narrator like Aria for historical anecdotes creates an inviting atmosphere, while the neutral Atlas voice works best for straightforward exhibit facts. This alignment helps visitors absorb information without feeling distracted by a mismatched tone.
Match tone to subject matter
Choose a voice that complements the subject. Use Sage for instructional content or Orion for serious historical sites. Avoid pairing upbeat Nova with solemn memorials to maintain context.
- Aria for warm historical narratives
- Atlas for neutral exhibit descriptions
- Orion for deep, documentary-style history
Consider audience demographics
Match the voice’s energy to your typical visitor. Younger crowds might appreciate Nova’s bright energy, while older visitors often prefer the clear, steady pace of Luna or Sage for better comprehension.
- Nova for energetic, younger audiences
- Sage for clear, instructional guidance
- Luna for soft, storytelling experiences
Test voices with sample text
Use your free trial credit to generate short clips of your actual guide script. Listen to how each preset handles your specific phrasing, ensuring the delivery matches your museum’s brand voice before committing.
- Generate samples using your guide script
- Listen for clarity and natural flow
- Compare options before finalizing
Language and accent control
Language control ensures your audio guides sound natural for international visitors. You can explicitly set the language using its ISO code, or let the system guess based on the text script. For non-English Latin-script text, we recommend setting the language field manually to avoid defaulting to English.
This precision matters because our 23 supported languages include German, French, Spanish, and many others. Explicit settings prevent mispronunciations that can distract tourists in museums or heritage sites.
Set language for accurate accent
Explicitly defining the language ensures the TTS engine applies the correct pronunciation rules and accent. This is crucial for multilingual museum tours where visitors expect clear, localized audio. Our API supports 23 languages, allowing you to serve accurate guides in Arabic, Chinese, Japanese, and more.
You can mix languages in your text by setting the language field appropriately. Each voice can speak any of these supported languages, ensuring consistent quality across different tour segments.
- Supports 23 languages including German, French, and Spanish
- Any voice can speak any supported language
- Explicit settings prevent mispronunciations
Auto-guess if not specified
If you do not specify a language, the system guesses based on the script. This works well for straightforward text but may fail with mixed-language content. For example, a German text with Latin script will be guessed correctly, but ambiguous phrases might default incorrectly.
We recommend always setting the language field for professional audio guides to ensure consistent delivery. This is especially important for historical sites where accuracy enhances the visitor experience.
- Guesses based on visible script
- May fail with mixed languages
- Recommended for simple text
Latin text defaults to English
If the language is not set and the text uses Latin characters, the system defaults to English. This can cause issues if your guide script is in another Latin-based language like Spanish or Italian. Always set the language to avoid unintended English delivery.
This default behavior simplifies basic use cases but requires attention for multilingual tours. Check your text script to ensure it matches the intended language setting.
- Defaults to English for Latin script
- Check text for mixed languages
- Set language for non-English Latin text

Testing voice options
Testing voice presets is straightforward with your free trial. You receive $0.50 of credit (about 50,000 characters) valid for 7 days upon signing in with Google. This allows you to generate sample audio without paying upfront, ensuring you find the perfect tone for your museum guides before committing to a top-up.
Use the API to create short clips for different exhibits. Compare how each voice handles factual descriptions versus storytelling sections. This practical approach helps you select the most engaging audio style for your specific audience and venue.
Use free trial credit
Sign up with Google to immediately access $0.50 in trial credit. This amount covers approximately 50,000 characters of text, which is sufficient to test multiple voice presets. The credit is valid for 7 days, giving you ample time to experiment with different tones without financial risk.
- - Sign up via Google for instant trial credit
- - Receive $0.50 valid for 7 days
- - Covers ~50,000 characters of text
Generate sample clips
Send your guide script snippets to the API endpoint. Request MP3 output for easy playback. Test various voices like Sage for instructions or Luna for stories. Each request takes a few seconds, allowing quick iteration. You can adjust speed settings between 0.25x and 4.0x to find the ideal pacing for your content.
Remember that longer texts are split into 4,096-character parts. For accurate testing, keep samples concise. Compare the audio files directly to see which voice resonates best with your visitors. This method ensures you choose the right tone before scaling up to full tour recordings.
- - Request MP3 output for easy playback
- - Adjust speed from 0.25x to 4.0x
- - Keep samples under 4,096 characters for simple testing
Listen on different devices
Play your generated audio clips on various devices typical for museum use. Test on smartphone speakers, tablet displays, and dedicated audio guide hardware. This ensures the voice clarity and tone translate well across different playback environments. Pay attention to how the voice sounds in noisy gallery settings versus quiet halls.
Listen to different voices in context. A voice that sounds clear on headphones might sound different on a small speaker. This step helps you avoid surprises when visitors actually use your audio guides. Ensure the selected voice remains engaging and understandable in real-world conditions.
- - Test on smartphones, tablets, and guide hardware
- - Check clarity in noisy vs. quiet environments
- - Verify tone translates well to small speakers
Consistency across tours
Using a single voice preset across all your audio guides establishes a consistent tone for your museum or tour service. When visitors hear the same distinctive tone in every exhibit, it creates a cohesive narrative thread that ties the entire experience together, reinforcing trust and professionalism without needing visual cues.
Consistency also reduces cognitive load for listeners. When the auditory style remains predictable, users can focus entirely on the historical facts or stories rather than adjusting to new vocal personalities with each turn. This subtle continuity enhances engagement and makes the educational content feel more polished and intentional.
Use same voice for branding
Assigning one voice to your entire collection helps build a sonic logo. Whether you choose the warm narration of Aria or the neutral news style of Atlas, maintaining that choice signals that the audio is officially part of your institution’s voice.
This approach works particularly well for large venues with multiple rooms. A unified sound ensures that guests perceive the audio as a single, integrated service rather than a collection of disparate recordings.
Create consistent user experience
Listeners appreciate predictability. When the tone matches across all exhibits, the transition between different topics feels natural. This consistency supports the educational goal by keeping the focus on the content rather than the delivery style.
For museums with diverse content, picking a versatile voice like Sage for instructional content ensures clarity without sacrificing warmth, keeping the audience engaged throughout their visit.
Avoid mixing voices randomly
Switching voices between adjacent exhibits can be jarring and distract from the historical narrative. It may feel like the audio was produced by different teams or agencies, diluting the professional feel.
If you must use different voices for vastly different content types, do so intentionally and clearly label the change. However, for most tourism applications, sticking to one preset yields the most cohesive result.
Random mixing can confuse visitors, making it harder for them to distinguish between official narration and separate informational clips.
- Stick to one preset per venue or tour line
- Use voice changes only for major thematic shifts
- Label new sections clearly if switching is necessary
Voice limitations to know
Our voices are designed for clear, natural speech, but they lack the dynamic emotional range of human performers. You cannot tag specific emotions like excitement or sadness to change the delivery style mid-sentence.
Additionally, we do not offer custom voice cloning. You cannot upload a sample of a curator’s voice to create a personalized clone; the eight presets are fixed and based on licensed recordings from LibriTTS-R and LibriVox.
No emotion tags available
The TTS engine does not support emotion tags. Each voice has a consistent emotional tone throughout the generation. For example, Aria remains warm and clear, while Felix stays energetic. You cannot instruct the AI to sound 'happy' or 'serious' in specific segments.
This consistency is often beneficial for museum guides, where a steady, predictable tone helps listeners focus on historical facts rather than fluctuating performance styles.
- No emotion tags supported
- Consistent tone per voice
- No dynamic pitch variation
No custom voice cloning
We do not offer custom voice cloning for customers. You cannot upload your own audio to create a unique voice. The eight presets (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion) are the only options available.
This ensures high quality and consistency across all users without the variability that comes with training custom models on limited samples.
Fixed voice characteristics
Each voice has fixed characteristics based on its licensed source recordings. While you can adjust speed and language, you cannot alter the fundamental timbre or accent of a voice.
This means a voice chosen for its clarity in a noisy museum hall will maintain that clarity regardless of the text. It also means you cannot mimic specific celebrity or politician voices.
Speed settings impact
Adjust speed from 0.25 to 4.0
Speed is a simple numeric setting between 0.25 and 4.0. It scales the voice’s natural pace without changing pitch or tone. This gives you precise control over how quickly information is delivered to museum visitors.
Slower speed for clarity
Set the speed lower for complex historical facts or non-native speakers. Slower delivery ensures every word is distinct, reducing the cognitive load for listeners navigating a noisy gallery. This is ideal for detailed exhibit descriptions.
Faster speed for brevity
Increase speed for short transitions or well-known points where brevity matters. Faster generation keeps the tour moving efficiently, preventing audio from dragging during simple wayfinding instructions or brief interludes between major exhibits.
Final selection tips
Selecting a voice preset for your audio guide is less about finding a single "best" option and more about aligning the tone with the specific content you are presenting. Whether you are describing a static exhibit or telling a historical story, the voice should feel natural to the subject matter.
Choose based on content type
Match the voice personality to the information density. Use neutral presets like Atlas or Sage for factual, instructional content where clarity is key. Reserve engaging voices like Aria or Luna for narrative sections to keep visitors interested.
For introductory clips or promotional highlights, energetic voices like Felix can capture attention quickly. Conversely, use deep voices like Orion for serious historical contexts to add authority.
- Use Atlas for clear, news-style facts
- Try Luna for soft storytelling
- Pick Felix for high-energy promos
Consider listener preferences
Test your audio on different devices to ensure the tone resonates with your audience. Some visitors may prefer the calm instruction of Sage, while others might enjoy the warm narration of Aria.
Since requests are processed sequentially, test your sample texts to ensure the voice remains clear even at slower speeds, which can improve comprehension for non-native speakers.
Stick to one voice per tour
Consistency builds a stronger brand experience. Avoid switching voices randomly between exhibits, as this can confuse visitors and break immersion.
Select one preset for an entire tour or exhibit path. This helps listeners form a connection with the guide character over time.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Support your international visitors with clear, natural audio in their native tongue.
23 languages
Any voice speaks any of these languages.
