Frequently asked questions
Which voice is best for VTubers?
There is no single best voice; choose based on your character’s persona. Aria offers a warm narrator style, while Luna provides soft storytelling. For energetic streamers, Felix or Nova work well. Atlas delivers a neutral news tone, and Orion suits deep documentary styles. Select the preset that matches your on-screen avatar’s personality and content type.
Can I mix voices in one video?
Yes. Generate each line or segment using the API with the desired voice preset. You can then edit these audio files together in your video editor. Since each request produces a separate file, you have full control over pacing and transitions between different voices within your final production.
Do the voices sound robotic?
The voices are based on licensed recordings from LibriTTS-R and LibriVox public-domain audiobook readings. However, every generated file carries an inaudible watermark marking it as synthetic speech, distinguishing it from human-recorded audio without affecting the audible quality.
Can I adjust the speed?
Yes. The API accepts a speed parameter ranging from 0.25 to 4.0. This allows you to slow down narration for clarity or speed up dialogue for energetic content. Adjust this value in your JSON request body alongside the model, input text, and voice selection parameters.
Are the voices gender-specific?
The eight presets are categorized as female (Aria, Nova, Luna, Sage) or male (Felix, Milo, Atlas, Orion). However, any voice can speak any of the 23 supported languages. You can assign any voice to any character regardless of traditional gender associations, allowing flexible casting for your VTuber projects.
Can I use the voices in YouTube videos?
Yes. The voices are derived from public-domain audiobook readings (LibriVox) and licensed recordings. You can use the generated audio in YouTube videos, streams, and other content. Since the input text is not used for training and audio is returned in the response, you retain the assets for your productions.
Do I need to credit the voices?
The source material comes from LibriVox public-domain readings and licensed recordings. The API does not mandate specific attribution for usage. Each file includes an inaudible synthetic watermark. You can use the voices in your content without mandatory credit, though acknowledging the source is always appreciated by the community.
Are the voices available in all languages?
Each of the 8 voices can speak any of the 23 supported languages, including English, Japanese, German, and Spanish. You can specify the language via the language parameter or let the API guess it from the script. Note that Vietnamese, Indonesian, and Thai are not supported.
Can I clone a specific voice?
No. We do not offer custom voice cloning or uploading of your own voice. You are limited to the 8 preset voices: Aria, Nova, Luna, Sage, Felix, Milo, Atlas, and Orion. These are fixed presets based on licensed recordings, providing consistent quality without the need for custom training.
Is there a free trial?
Yes. Signing in with Google provides $0.50 of credit, valid for 7 days, allowing about 50,000 characters. Email/password accounts start without trial credit. You can top up using USDT on the TRON network. Failed requests are free, so you can test endpoints without incurring charges.
How do I choose a voice?
Review the eight presets: Aria (warm narrator), Nova (bright upbeat), Luna (soft storytelling), Sage (calm instructional), Felix (energetic promo), Milo (friendly conversational), Atlas (neutral news), and Orion (deep documentary). Test them with your specific text using the API. The best choice depends on the emotional tone of your VTuber persona.
Can I use the voices for ads?
Yes. The API is suitable for commercial use, including advertisements for streaming services, games, or products. Since the source material is public-domain or licensed, and you receive the generated audio file, you can incorporate it into promotional content. There are no additional restrictions on commercial usage beyond the API terms.
Do the voices have accents?
The voices are designed to speak standard English or the specified language naturally. While they may have subtle tonal characteristics (e.g., Aria is warm, Felix is energetic), they do not have strong regional accents unless specified by the language parameter. The language field helps ensure correct pronunciation for non-English texts.
Can I use the voices for podcasts?
Yes. The API supports generating audio files in formats like MP3, WAV, and FLAC, which are standard for podcasting. You can generate episodes using any of the 8 voices. Since requests are processed sequentially, plan your generation workflow to avoid queue delays during high-volume production periods for your podcast series.
Female Voices Overview
VocalWave provides four distinct female voice presets, each engineered for specific content styles. These voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and from LibriVox public-domain audiobook readings.
Select the appropriate voice based on your avatar’s personality and the emotional tone of your video. Each preset offers a unique acoustic profile, ensuring consistent audio assets for your channel.
Aria: Warm Narrator
Aria delivers a warm, clear narration style ideal for storytelling or explanatory content. Her voice provides a natural, engaging flow that keeps viewers attentive without sounding artificial.
- Ideal for lore videos or character backstories
- Clear enunciation for complex narratives
Nova: Bright and Upbeat
Nova offers a bright, energetic tone perfect for high-energy streams or quick updates. Her upbeat delivery matches the fast-paced nature of gaming highlights or reaction content.
- Best for dynamic gaming segments
- Engages younger audiences effectively
Luna: Soft Storyteller
Luna provides a soft, soothing voice suitable for calming content, ASMR-adjacent videos, or gentle introductions. Her tone reduces listener fatigue during long-form content.
- Suitable for relaxing gameplay streams
- Soft tone for evening content
Sage: Calm Instructional
Sage is designed for instructional content, tutorials, or neutral commentary. Her calm demeanor ensures clarity and professionalism, making complex information easier to digest.
- Perfect for tutorial videos
- Neutral tone for reviews
Male Voices Overview
The male presets cover distinct broadcast styles, from high-energy promo to deep documentary narration. Each voice is engineered for clarity and consistency, ensuring your VTuber avatar’s audio remains professional without manual tuning.
Felix: Energetic Promo
Felix delivers high-impact energy, ideal for promotional segments or dynamic stream intros. His tone cuts through background noise, making him suitable for hype moments or quick announcements.
Milo: Friendly Conversational
Milo offers a warm, approachable sound perfect for direct address or chat interaction. His conversational rhythm feels natural, helping maintain viewer engagement during long streams.
Atlas: Neutral News
Atlas provides a clean, impartial voice ideal for factual updates or neutral commentary. His consistent tone ensures information is delivered clearly without emotional bias.
Orion: Deep Documentary
Orion’s deep, resonant quality suits storytelling or atmospheric segments. His voice adds gravity and depth, enhancing narrative-driven content or lore explanations.
Choosing for VTuber Content
Selecting the right voice preset is critical for maintaining character consistency across your VTuber content. The eight available presets—ranging from Aria's warm narration to Felix's energetic promo style—allow you to align audio tone with your avatar's visual personality.
Proper voice selection enhances emotional resonance and helps retain audience attention by ensuring the spoken content matches the on-screen persona's energy levels without requiring complex voice acting or custom training.
Avatar Matching
Match your avatar's visual design with the appropriate voice preset. For example, use Luna for soft, storytelling avatars or Orion for deep, documentary-style characters.
Avoid mismatching high-energy voices with calm visuals or vice versa. Consistency between visual and audio cues reduces cognitive load for viewers and strengthens character recognition over time.
- Use Luna for soft, storytelling avatars
- Use Orion for deep, documentary-style characters
- Match energy levels to visual design
Emotional Tone
Each preset carries a distinct emotional quality. Sage offers calm instructional tones, while Nova provides bright, upbeat energy. Choose based on the content type—educational segments benefit from Sage, while dynamic gameplay moments suit Nova.
Avoid overusing emotional extremes. Consistent tone helps viewers predict and engage with your content, reducing fatigue from constant high-energy delivery.
- Sage for calm, instructional segments
- Nova for bright, upbeat energy
- Use Luna for soft storytelling
Audience Retention
Consistent voice presets improve audience retention by creating a recognizable audio signature.
Test different presets with your audience to identify which tones resonate best. Use feedback to refine voice selection for future content, ensuring long-term engagement and loyalty.
- Create recognizable audio signatures
- Test presets with audience feedback
- Refine voice selection for engagement
Choosing for Stream Intros
Stream intros demand immediate impact. VocalWave’s API lets you generate consistent, high-quality audio assets quickly, ensuring your channel’s identity remains recognizable across episodes.
Energy Levels
Match Felix’s energetic promo style for high-octane gaming streams, or use Milo’s conversational tone for chill streams. Adjust speed between 0.25–4.0 to fit your pacing.
- Use Felix for energetic intros
- Use Milo for conversational segments
- Adjust speed to match content rhythm
Brand Identity
Consistency builds recognition. Stick to one or two voices for your intro theme. This creates an auditory logo that viewers associate with your brand.
Audio Clarity
Clear audio is critical for short clips. Use mp3 or wav formats. The 8 presets are designed for clarity.
- Use mp3 or wav formats
- Optimized for clarity
- Minimal editing required
Language Support
VocalWave supports 23 languages, covering major markets for VTubers and streamers. The system handles voice generation for Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish.
23 Languages Available
Every voice preset can speak any of the 23 supported languages. This flexibility allows you to produce consistent character voices across different linguistic audiences without training new models. Unsupported languages include Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, and Bengali.
Default Language Detection
If you do not specify a language, the API guesses it from the text script. Latin script defaults to English. This works well for standard English content but may fail for other languages using Latin characters.
Manual Override
For non-English Latin-script text, explicitly set the language field. This prevents mispronunciation and ensures the voice uses the correct phonetic rules. Always define the language parameter for multi-lingual streams or mixed-language scripts.
Technical Considerations
VocalWave’s TTS API offers precise control over audio output, allowing VTubers to tailor generated speech to their specific workflow needs. By adjusting speed and selecting the appropriate audio format, creators can optimize assets for streaming overlays, video edits, or live commentary without compromising quality.
The API supports speed adjustments from 0.25x to 4.0x, enabling fine-tuning of pacing for different content types. Audio can be exported in multiple formats to suit various editing software and streaming platforms, ensuring seamless integration into existing production pipelines.
Speed Adjustment
Adjust the speed parameter between 0.25 and 4.0 to control the pacing of the synthesized voice. Slower speeds (e.g., 0.75x) are ideal for clear, instructional content or detailed lore explanations, while faster speeds (e.g., 1.25x–1.5x) suit energetic stream intros or quick updates.
Speed changes do not affect character pricing; you are charged based on the original character count of the input text. This allows you to experiment with pacing without incurring additional costs for re-generating the same script.
- Range: 0.25 (slow) to 4.0 (fast)
- Default speed is 1.0x
- No extra cost for speed adjustments
Format Selection
Choose from MP3, WAV, Opus, FLAC, or PCM via the response_format field. MP3 is recommended for general streaming and video editing. WAV or FLAC are better for high-fidelity post-production where lossless quality is preferred.
PCM provides raw audio data, which can be useful for specific software integrations that require uncompressed input. Opus is efficient for web-based playback with low bandwidth requirements.
- MP3: Best for streaming and video editing
- WAV/FLAC: Lossless quality for post-production
- PCM: Raw audio for specific integrations
- Opus: Efficient for web playback
Character Limits
Each API request supports up to 4,096 characters of input text. For longer scripts, split the text into multiple requests and join the resulting audio files. This ensures consistent quality and avoids potential processing timeouts.
Spaces and punctuation count toward the character limit. Plan your script length accordingly, keeping individual requests under the 4,096-character threshold for optimal performance.
- Max 4,096 characters per request
- Spaces and punctuation count as characters
- Split long texts into multiple requests
Comparison with Custom Cloning
Custom voice cloning requires recording hours of reference audio, training a model, and often iterative tuning to achieve a usable result. Our approach offers eight pre-trained presets based on licensed recordings from LibriTTS-R and LibriVox, providing immediate consistency without the overhead of custom training pipelines.
For VTubers needing reliable, license-safe assets, presets deliver uniform quality across all content types. Cloning introduces variability and higher costs for marginal voice-specific nuance that most streaming audiences cannot distinguish from well-chosen presets.
Cost Difference
Cloning services typically charge setup fees ranging from $50 to $200 per voice. With VocalWave, you pay a flat $0.01 per 1,000 characters. There are no hidden training fees, no per-voice subscriptions, and no minimum spend requirements.
- No upfront cloning fee
- Pay only for characters generated
- Transparent character-based billing
Quality Differences
Custom clones often retain minor artifacts or inconsistent tone across different phonemes. Our presets are optimized for clarity and consistency. While a clone matches a specific timbre, presets offer broader emotional range, such as Luna’s soft storytelling or Felix’s energetic promo style, which a static clone may lack.
- Consistent tone across phrases
- Distinct emotional characteristics per voice
- No training artifacts
Setup Time
Cloning can take days to weeks for high-fidelity models. With VocalWave, you select a voice and start generating audio immediately after signing up. Your first request works instantly, with no waiting for model training or quality assurance checks.
- Immediate availability
- No training wait time
- Instant API integration

Licensing and Usage
VocalWave utilizes voice models derived from the LibriTTS-R dataset, which is licensed under CC BY 4.0. This foundation provides a clear legal framework for the synthetic nature of the audio files generated by our API.
CC BY 4.0 License
The core voice performances originate from LibriVox public-domain audiobook readings, standardized under the Creative Commons Attribution 4.0 International license. This provides a clear legal framework for the synthetic nature of the audio files generated by our API.
Commercial Use Allowed
You may use the generated audio files for commercial purposes, including monetized VTuber streams, YouTube videos, and podcast productions. There are no restrictions on the volume of content you create, provided you have sufficient API credit.
No Attribution Required
While the source data is CC BY 4.0, you are not required to attribute VocalWave or the original LibriVox readers in your video descriptions or stream overlays. This allows for a clean, professional presentation of your content.
Performance Metrics
Understanding generation speed and latency helps VTubers plan their live streams and video edits accurately. The system processes requests sequentially, meaning performance depends on current load and text length rather than parallel processing power.
With a typical generation speed of roughly 25 characters per second, long scripts take time to synthesize. Short phrases may take a few seconds, while extended monologues approach real-time playback duration.
Generation Speed
Text-to-speech conversion averages 25 characters of text per second of generated audio. This means a 10,000-character script will take approximately 6.5 minutes to produce. Longer texts are split into 4096-character chunks and joined seamlessly.
- Approx. 25 chars/sec generation rate
- Real-time speed for long texts
- Short phrases take a few seconds
Latency
Latency varies based on server load. Since requests are processed one after another, higher traffic increases wait times. The API does not guarantee sub-second responses for large inputs.
Queue Times
Under heavy load, requests may wait in a queue before processing begins. This is standard for shared infrastructure. Ensure your production workflow accounts for potential delays during peak hours.
Pricing Impact
VocalWave offers two model variants, tts-1 and tts-1-hd, which use the same underlying engine and incur identical pricing. There is no cost difference between high-definition and standard quality tiers, simplifying your budget calculations for VTuber assets.
tts-1 vs tts-1-hd
Both models process text through the same synthesis pipeline, ensuring consistent voice characteristics across presets. You can switch between tts-1 and tts-1-hd without penalty, allowing you to test output quality for specific content types without adjusting your expected costs.
Cost per Minute
Pricing is strictly character-based at $0.01 per 1,000 characters. Since generation speed averages roughly real-time, a one-minute audio file typically consumes around 150 characters of input text, costing approximately $0.0015.
Budget Planning
Start with the $0.50 Google Sign-In trial (50,000 characters) to prototype your VTuber scripts. For larger projects, top up via USDT (TRC-20) with a $5 minimum. Failed requests are refunded, so you only pay for successfully generated audio files.
- Trial: $0.50 credit for 7 days
- Standard rate: $10 per 1M characters
- Top-up: USDT (TRC-20), min $5
FAQs
VTuber creators often ask about voice selection and technical constraints when integrating TTS into their workflows. Understanding the capabilities and limitations of preset voices helps streamline content production without requiring custom training or complex setup processes.
Which voice is most popular?
Popularity varies by content style. Aria and Nova are frequently chosen for narrative and upbeat segments, while Atlas suits news-style updates. There is no single dominant voice; selection depends on the specific tone required for each video or stream segment.
Can I change voices mid-stream?
The API generates audio files per request, so you cannot switch voices in real-time within a single audio stream. To change voices, you must generate separate audio files for each voice segment and concatenate them during editing or playback.
Do voices sound generic?
Voices are based on licensed recordings from LibriVox and LibriTTS-R. While they are distinct presets, they may lack the unique character of celebrity clones. Each file includes an inaudible watermark identifying it as synthetic speech, which may be detected by some platforms.
Can I use multiple voices?
Yes. You can utilize any of the eight preset voices (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion) across your projects. Each voice can be assigned to different characters or segments, provided you generate separate audio files for each.
Getting Started
Create your account via /get-api-key/ using Google or email. You receive immediate access to the dashboard, where your API key and balance are displayed.
Generate your first key instantly. Use it with any OpenAI-compatible SDK by pointing the base URL to https://api.playbet177.com/v1.
Sign Up Process
Visit /get-api-key/ and choose Google or email registration. Accounts are site-specific. Google sign-ups receive a $0.50 trial credit valid for 7 days. Email accounts start with zero balance.
- Google or email registration
- $0.50 trial for Google users
- Instant dashboard access
API Key Generation
Your key appears immediately on the dashboard. It starts with 'tts-'. One active key per account; issuing a new one invalidates the old one instantly.
- Starts with 'tts-'
- One active key per account
- Old keys invalidate immediately
First Request Example
Send a POST to /v1/audio/speech with model 'tts-1' or 'tts-1-hd', your text input, and a voice preset like 'aria'. Receive an MP3 file directly in the response.
- Endpoint: POST /v1/audio/speech
- Model: tts-1 or tts-1-hd
- Response: Direct audio file
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Your first request works instantly, with no waiting for model training or quality assurance checks.
23 languages
Any voice speaks any of these languages.
