Frequently asked questions
Which voice is best for kids?
Nova is bright and upbeat, making it ideal for engaging young learners. Aria offers a warm, clear narrator tone suitable for educational content. Both voices support English and other listed languages, allowing you to create distinct, character-based voices for different course segments without hiring voice actors.
Can I mix Aria and Felix?
Yes, you can mix voices by splitting your text into parts. Send separate requests for Aria and Felix, then join the resulting audio files. Each request supports up to 4096 characters. This workflow lets you create dynamic game show host styles or varied narrator tones for your LMS modules.
Do voices sound different in MP3 vs WAV?
The voice performance is identical across formats; only the file container changes. You can choose MP3, WAV, OPUS, FLAC, or PCM based on your LMS compatibility needs. MP3 offers smaller file sizes, while WAV provides uncompressed audio. All formats are generated via the same engine with consistent quality.
Is Nova too energetic for some courses?
Nova is described as bright and upbeat. If this energy level feels too high for your specific course material, you can adjust the speed parameter between 0.25 and 4.0. Slowing down Nova’s delivery can make it more suitable for instructional content while retaining its clear, engaging character.
Which voice is best for English learners?
Aria provides a warm, clear narrator tone, which is excellent for language learning. Sage offers a calm, instructional style suitable for clear pronunciation guides. Both voices support English and other languages, ensuring consistent clarity for learners. You can adjust speed to match the learner’s proficiency level.
Can I use Orion for news updates?
Yes, Orion is described as deep and documentary-style, making it well-suited for news updates. Its tone adds authority to instructional content. You can generate these updates via the API, specifying the Orion voice and adjusting speed if needed to match the pacing required for your online course modules.
Do voices work well in background?
Voices are designed for clear narration. For background use, you can adjust the speed parameter (0.25–4.0) to fit your audio mix. The audio is generated as standard files (MP3, WAV, etc.) that you can integrate into your LMS. There is no official streaming, so you manage playback yourself.
Is Sage too formal for games?
Sage is calm and instructional, which may feel too formal for energetic game shows. For games, consider Felix (energetic, promo) or Nova (bright, upbeat). You can mix voices to create varied host styles. Sage is better suited for clear instructions or calm segments within your course content.
Can I change voice speed per request?
Yes, you can adjust the speed parameter for each request between 0.25 and 4.0. This allows you to fine-tune the pacing of game show hosts or narrators. Adjusting speed helps match the tone to your specific course content without needing to regenerate audio with a different voice preset.
Which voice is most natural sounding?
All voices are based on licensed recordings from LibriTTS-R and LibriVox. Each voice has a distinct character: Aria is warm, Felix is energetic, and so on. Naturalness depends on your preference for tone. There is no single "most natural" voice; each preset offers a different, realistic style for your needs.
Do voices support humor well?
Felix is energetic and promo-style, which may suit humorous or lively content. Nova is bright and upbeat, also suitable for engaging, light-hearted segments. You can adjust speed to enhance comedic timing. While there are no emotion tags, the distinct character of each voice helps convey different tones effectively.
Can I use Milo for customer support?
Milo is friendly and conversational, making it suitable for customer support-style audio in courses. You can generate these lines via the API. Each request supports up to 4096 characters. You can adjust speed to ensure clarity, ensuring the tone remains helpful and approachable for your learners.
Is Atlas too monotone?
Atlas is described as neutral and news-oriented. It may sound less expressive than Felix or Nova. If you find it too monotone, you can increase the speed slightly or switch to a more energetic voice like Felix. Atlas is best for clear, straightforward information delivery in your LMS.
Which voice is best for instructions?
Sage is calm and instructional, making it ideal for clear, step-by-step guidance. Aria is also a strong choice with its warm, clear narrator tone. Both voices ensure clarity for learners. You can adjust speed to match the complexity of the instructions, ensuring the audio is easy to follow.
Can I use Luna for bedtime stories?
Luna is soft and storytelling-oriented, making it perfect for bedtime stories or calm narrative segments. You can generate these using the API, specifying the Luna voice. Adjusting the speed to a slower pace can enhance the soothing tone, creating an immersive experience for your online course content.
Do voices work in noisy environments?
Voices are generated as standard audio files (MP3, WAV, etc.) that you can control in your LMS player. Clarity depends on your playback volume and environment. You can adjust speed or choose a clear voice like Aria for better intelligibility. There is no noise cancellation built into the API.
Aria: Warm and clear narrator
Aria delivers a warm, clear narration that serves as an ideal foundation for instructional guides. Her steady pace and natural intonation help learners absorb new concepts without distraction, making complex information feel approachable and easy to follow.
Ideal for instructional guides
When introducing new topics, Aria’s consistent tone keeps learners focused on the material rather than the voice itself. This clarity reduces cognitive load, allowing students to process instructions more effectively.
Clear enunciation for details
Technical terms and specific instructions are articulated with precision. Aria’s crisp enunciation ensures that key details—such as software steps or safety protocols—are heard clearly, reducing the need for learners to re-listen to segments.
Best for step-by-step modules
For sequential tasks, Aria maintains a rhythm that aligns well with logical progression. Her voice guides learners through each phase smoothly, ensuring no step is rushed or muddled by excessive energy or emotion.
Nova: Bright and upbeat
Nova brings a bright, upbeat energy to your LMS modules, making her ideal for keeping learners engaged during quick knowledge checks. Her clear, energetic tone transforms standard quizzes into lively interactions without feeling forced.
For courses targeting younger audiences or requiring a lively atmosphere, Nova’s voice adds a layer of enthusiasm that encourages participation. She is particularly effective in high-energy segments where you need to maintain momentum and interest.
Perfect for energetic quizzes
Nova’s upbeat delivery works best for rapid-fire quiz questions or quick feedback segments. Her clarity ensures that instructions are understood immediately, reducing cognitive load for learners while keeping the pace brisk and engaging.
Engaging for younger audiences
When teaching younger learners, a warm yet energetic tone helps maintain attention. Nova’s style feels approachable and lively, making complex topics feel less like a lecture and more like an interactive conversation.
Best for high-energy segments
Use Nova for game show host roles or competitive challenges within your course. Her energetic voice style supports fast-paced activities, ensuring that the audio matches the visual excitement of the on-screen content.
Luna: Soft storytelling voice
Luna’s soft, storytelling voice transforms dense instructional material into an engaging narrative experience. Her gentle cadence helps learners absorb complex information without feeling overwhelmed, making her an ideal companion for deep-dive modules where clarity and comfort matter more than high-energy excitement.
Unlike the bright enthusiasm of Nova or the calm tone of Sage, Luna provides a soothing backdrop that encourages sustained attention. This makes her particularly effective for guiding users through multi-step processes or detailed explanations where a lighter touch prevents cognitive fatigue.
Calming for complex topics
When introducing intricate concepts, Luna’s steady pacing allows learners to process information at a comfortable rhythm. Her voice avoids the energy of Felix or the neutrality of Atlas, creating a comfortable space for understanding difficult material.
The soft timbre reduces the perceived difficulty of the content, helping students feel more confident as they navigate through challenging subjects. This is especially useful for onboarding modules or compliance training where retention is key.
- Reduces cognitive load during dense explanations
- Gentle pacing aids information retention
- Ideal for reducing anxiety in complex topics
Ideal for narrative modules
Luna excels at delivering story-based learning content. Her natural storytelling rhythm keeps learners engaged without demanding high energy, making her perfect for case studies or historical overviews in your LMS.
She bridges the gap between educational content and entertainment, ensuring that narrative-driven lessons feel immersive rather than mechanical. This approach helps maintain interest throughout longer course segments.
By using Luna for narrative sections, you create a consistent emotional tone that supports the learning objectives without distracting with excessive personality.
Best for reflective learning
Reflective learning moments benefit from Luna’s calming presence. Her voice encourages pause and consideration, making her suitable for quiz intros, summary recaps, or guided meditation-style learning breaks.
The soft delivery invites learners to think critically rather than react quickly. This aligns well with adult learning principles that value thoughtful engagement over rapid-fire information delivery.
Use Luna to signal transitions between high-energy activities and deeper reflection, helping students mentally prepare for the next phase of their learning journey.
Sage: Calm and instructional
Sage delivers a calm, instructional tone ideal for formal training modules. Its steady pacing reduces cognitive load, helping learners absorb complex procedures without distraction. This voice prioritizes clarity over flair, ensuring every instruction is heard distinctly.
Use Sage for professional courses requiring a neutral, reliable delivery. It works well for compliance training, technical manuals, or detailed process guides where consistency matters more than emotional engagement.
Authoritative yet gentle
Sage balances authority with approachability. It sounds professional without being stiff, making it suitable for structured learning environments.
The voice maintains a consistent rhythm, reducing listener fatigue during longer instructional segments. This stability helps learners focus on content rather than voice delivery.
- Ideal for step-by-step procedural guides
- Reduces cognitive load in complex topics
- Maintains professional tone without rigidity
Best for formal training
Sage excels in formal settings where clarity is paramount. Its neutral tone avoids distracting listeners with unnecessary emotional cues.
Use this voice for compliance modules, policy explanations, or technical onboarding where precision and reliability are key to effective learning.
Ideal for professional courses
For professional development courses, Sage provides a steady, trustworthy delivery. It supports detailed explanations without overwhelming the learner.
Its calm nature ensures that complex information is processed effectively, making it a reliable choice for adult learners in corporate or academic settings.
Felix: Energetic promo style
Felix brings a bright, promotional energy to your LMS content, making him ideal for high-stakes quiz segments or competitive learning modules. His clear, upbeat delivery keeps learners engaged during fast-paced challenges.
High energy for competitions
Felix’s tone is designed to match the excitement of game show hosts, ensuring learners stay alert during timed quizzes. His energetic delivery works well for leaderboards and instant results.
Best for game show hosts
Use Felix to simulate a host announcing questions or prizes. His energetic style adds a layer of excitement without hiring voice actors for short instructional clips.
Ideal for quick feedback
Generate concise audio feedback for correct or incorrect answers quickly. Felix’s clear enunciation ensures instructions are understood immediately, reducing cognitive load for learners in rapid-fire scenarios.
Milo: Friendly and conversational
Milo brings a friendly, conversational tone to your LMS courses, making complex concepts feel like a chat with a knowledgeable peer. His approachable voice reduces learner anxiety and encourages engagement during interactive segments.
Unlike formal narrators, Milo sounds natural and relatable, which is crucial for keeping students attentive during long modules or feedback loops.
Approachable tone for learners
Milo’s voice strikes a balance between professional and casual, creating a comfortable learning environment. This tone helps learners absorb information without feeling lectured, fostering a sense of connection with the course material.
- Reduces cognitive load through familiar speech patterns
- Encourages repeated listening without fatigue
- Ideal for introductory modules
Best for interactive dialogues
Use Milo for quiz feedback, scenario-based learning, or conversational practice. His natural cadence mimics real human interaction, making scripted dialogues feel less robotic and more engaging for students practicing language skills or soft skills.
- Perfect for role-play exercises
- Supports conversational practice in foreign languages
- Enhances scenario-based training modules
Ideal for peer-like guidance
Milo works well as a ‘study buddy’ voice for study tips, reminders, or gentle nudges. His tone feels supportive rather than authoritative, making it ideal for encouraging learners to complete tasks or review material.
- Great for progress reminders
- Suitable for motivational messages
- Works well for mentor-style content
Atlas: Neutral news announcer
Atlas provides a neutral voice suitable for news-style updates. It delivers information with clarity and consistency, making it ideal for LMS modules where the focus is on the content rather than the personality of the speaker. This voice minimizes distraction, allowing learners to concentrate on facts and data.
Professional and clear
Atlas speaks with a steady, measured cadence that avoids emotional bias. This neutrality ensures that instructional material is presented objectively, which is crucial for compliance training or policy updates where tone should not influence interpretation of the rules.
- Consistent pacing across long texts
- No emotional inflection to distract from data
- Clear enunciation for complex terminology
Best for factual content
When your course relies on statistics, dates, or procedural steps, Atlas ensures these details are delivered accurately. Its straightforward style is reliable for presenting essential information without adding unnecessary narrative flair.
- Ideal for compliance and regulatory updates
- Clear delivery of numerical data
- Reduces cognitive load for dense text
Ideal for news-style updates
Use Atlas for weekly course updates or breaking news within your learning platform. Its broadcast-style tone helps learners recognize these segments as distinct from core lessons, signaling that the content is current, time-sensitive, and meant for quick consumption before moving to deeper study.
- Signals time-sensitive content clearly
- Separates updates from core curriculum
- Professional tone for organizational announcements

Orion: Deep documentary voice
Orion brings a deep, resonant tone that adds weight to your LMS content. His voice cuts through background noise clearly, ensuring learners stay focused on the material without distraction. This steady delivery works well for complex subjects that require careful attention.
Authoritative and deep
The low frequency of Orion’s voice commands attention naturally. It provides a professional backdrop that feels serious without being intimidating. This depth helps distinguish key takeaways from standard instructional text.
- Deep resonance reduces listener fatigue
- Clear projection for noisy environments
- Professional tone for corporate training
Best for serious topics
Use Orion when discussing compliance, safety protocols, or financial data. His steady pace ensures clarity without rushing through critical details. It signals to learners that this information matters.
- Safety compliance modules
- Regulatory updates
- Medical or legal briefings
Ideal for documentary style
Pair Orion with historical or scientific content to create a documentary feel. His narration style is suitable for documentary-style content.
- Historical case studies
- Scientific explanations
- Long-form narrative modules
Comparing voice styles for engagement
Choosing the right voice style directly impacts learner engagement in LMS game shows. By aligning vocal tone with module objectives, instructional designers can reduce cognitive load and maintain attention during high-stakes quiz segments or reflective learning exercises.
This section compares energy levels, formality, and audience preferences to help you select the most effective voice for your specific course structure.
Energy levels compared
Voices range from high-energy to calm. Felix and Nova deliver bright, upbeat tones ideal for quick feedback and energetic quizzes. In contrast, Luna and Sage offer softer, more relaxed delivery suitable for complex topics or professional training where a less intense tone aids comprehension.
- High Energy: Nova, Felix
- Medium Energy: Aria, Milo
- Low Energy: Luna, Sage, Atlas, Orion
Formality scale analysis
The formality spectrum allows you to match voice to content type. Atlas provides a neutral, news-announcer style perfect for factual updates. Sage offers an authoritative yet gentle tone for formal training. Milo remains friendly and conversational, ideal for peer-like guidance in interactive dialogues.
Mixing these tones helps distinguish between instructional content and game elements, keeping the learning experience dynamic without confusing the learner.
- Formal: Atlas, Sage
- Neutral: Aria, Milo
- Casual: Nova, Felix, Luna, Orion
Audience preferences
Different learner groups respond better to specific vocal characteristics. Younger audiences often prefer the bright, upbeat energy of Nova or Felix, which keeps them engaged during fast-paced segments. Professional learners may find Sage or Atlas more credible for serious training modules.
General audiences typically connect well with Aria or Milo, who offer a balanced, approachable tone suitable for mixed-content courses. Use Luna or Orion for reflective, documentary-style modules requiring deeper focus.
- Youth: Nova, Felix
- Professionals: Sage, Atlas
- General: Aria, Milo
- Mixed: Luna, Orion
Language compatibility for voices
All eight voices support the same 23 languages, allowing you to maintain consistent character voices across multilingual courses. Whether your learners speak Arabic, Chinese, or Swahili, you can assign the same host persona without losing tonal continuity.
You simply provide the ISO language code in your API request. The system handles the rest, ensuring your educational content remains accessible without hiring separate voice talent for each region.
All voices support 23 languages
Every voice preset—Aria, Nova, Luna, Sage, Felix, Milo, Atlas, and Orion—can speak all 23 supported languages. This uniformity ensures that your game show host character sounds identical regardless of the language being spoken.
You do not need to map specific voices to specific languages. A single voice model delivers consistent personality across English, Japanese, German, and the other 20 supported tongues.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
Native pronunciation quality
The voices are trained to produce native-level pronunciation for each supported language. This reduces the 'robotic' feel often associated with non-native TTS models, keeping learners engaged.
If you do not specify a language, the system guesses it from the script. For Latin-script text outside English, explicitly set the language parameter to avoid incorrect defaults.
Cross-language consistency
Since the same voice presets handle all languages, your course maintains a unified audio identity. Learners transitioning between language modules will recognize the same host tone and style.
This consistency supports brand recognition in your LMS without requiring complex voice management. Note that unsupported languages like Vietnamese or Ukrainian are not available.
Selecting voices for course structure
Structure your LMS modules by assigning distinct voices to different content types. Use Sage for formal training sections and Luna for reflective learning parts. This separation helps learners mentally categorize information, reducing cognitive load during long courses.
Avoid using the same voice for every segment. Mixing voices keeps learners engaged by providing auditory variety without changing the visual layout. Ensure the transition between voices feels natural within your course design.
Matching tone to module type
Align voice personality with module goals. Use Felix for high-energy quizzes and competition segments. Select Atlas for factual updates or news-style summaries. This match reinforces the content’s intent without requiring visual cues.
- Use Felix for quick feedback loops
- Use Sage for formal training
- Use Luna for narrative modules
- Use Atlas for factual content
Mixing voices for variety
Rotate voices across modules to maintain interest. Switch from Aria in instructional guides to Nova in energetic quizzes. This variety prevents auditory fatigue during multi-hour learning sessions.
- Rotate voices every 5-10 minutes
- Pair energetic voices with quizzes
- Use calm voices for reviews
Consistency across modules
Maintain the same voice for related modules. If a character appears in multiple lessons, keep their voice consistent. This builds recognition and trust with your learners over time.
Technical performance of voices
Technical performance varies slightly by voice model and format, but generation speed is generally consistent at roughly real-time for long texts. Short phrases may take a few seconds to process, especially under load when requests queue.
Audio quality remains high across all models, with tts-1-hd offering enhanced clarity for complex instructional content. File sizes depend on the selected format and speed, allowing you to optimize for bandwidth or playback compatibility.
Generation speed differences
Most voices generate audio at approximately 25 characters per second. Short inputs may experience a slight initial delay as the engine initializes, but processing is sequential.
Under heavy load, requests may queue, adding a few seconds to the total time. The tts-1-hd model processes at the same speed as tts-1.
- Roughly 25 characters of text per second of audio.
- Short phrases take a few seconds to generate.
- Requests are processed sequentially; queues may add delay.
Audio quality comparisons
The tts-1-hd model provides higher fidelity audio, suitable for detailed educational modules where clarity is paramount. The tts-1 model is optimized for speed with clear, natural speech.
All voices carry an inaudible synthetic watermark, ensuring content authenticity without affecting listener experience.
- tts-1: Standard quality, optimized for speed.
- tts-1-hd: Higher fidelity, better for complex text.
- Inaudible watermark included in all outputs.
File size variations
File size is determined by the chosen format and audio speed. MP3 files are smallest, ideal for web embedding. WAV and FLAC offer lossless quality for archiving.
Selecting a higher speed (e.g., 2.0x) reduces file size by shortening duration. PCM files are larger and uncompressed, best for post-production editing.
- MP3: Smallest size, good for web.
- WAV/FLAC: Lossless, larger files.
- PCM: Uncompressed, largest size.
- Speed affects duration and file size.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Technical terms and specific instructions are articulated with precision.
23 languages
Any voice speaks any of these languages.
