Frequently asked questions
Which voice is best for RPGs?
Q: Which voice is suitable for RPGs? A: Orion suits deep, documentary-style characters, while Luna offers soft storytelling for quests. Atlas provides neutral dialogue, and Sage works well for instructional elements. Aria adds a warm narrator touch. Since all eight presets support any language, choose based on your character’s personality rather than a single "best" option.
Can I mix male and female voices?
Yes, you can mix male and female voices freely. The API allows you to select any of the eight presets for each request. You can switch between voices like aria and felix within the same scene or project without any restrictions or additional fees.
Do I need to pay for each voice?
No, all eight voice presets are included in the standard pricing. You pay $0.01 per 1,000 characters regardless of which voice you choose. There are no extra fees for using specific voices like nova, milo, or atlas within your game dialogue.
Are the voices realistic?
The voices are based on licensed recordings from LibriTTS-R and LibriVox. They sound natural but carry an inaudible watermark marking them as synthetic speech. They are clear and distinct, suitable for NPC dialogue, but do not claim to be indistinguishable from human actors.
Can I change the voice mid-sentence?
No, each API request generates audio for a single text input using one voice. To change voices mid-sentence, you must split the text into separate requests, generate individual audio files, and join them in your game engine. The API does not support mid-sentence voice switching.
Do I need to download voice packs?
No, you do not need to download any voice packs. The voices are processed on our cloud server. You simply send your text via the API, and receive the audio file directly in your response. All processing happens remotely, keeping your game client lightweight.
Can I use these voices in mobile games?
Yes, the generated audio files can be used in mobile games. Since the processing happens in the cloud, your mobile device only needs to handle the final audio playback. This works well for both iOS and Android projects using standard audio formats like mp3 or wav.
What is the best format for web games?
Q: What format is suitable for web games? A: MP3 and OPUS are ideal for web games due to their small file sizes and broad browser compatibility. MP3 offers universal support, while OPUS provides efficient compression. Both formats ensure smooth audio loading and playback in modern browsers, making them reliable choices for NPC dialogue and sound effects in your projects.
Do the voices sound natural?
The voices are based on licensed recordings and sound natural for synthetic speech. They support 23 languages and can be adjusted for speed. However, they are not human recordings and include an inaudible watermark. They are designed for clear NPC dialogue rather than cinematic realism.
Can I adjust the pitch?
No, the API does not offer pitch adjustment. You can only adjust the speed between 0.25 and 4.0. To change the perceived pitch, you would need to use your game engine’s audio tools or post-processing effects after receiving the generated audio files.
Do I need an internet connection?
Yes, you need an internet connection to use the API. Since the text is sent to our cloud server for synthesis, your game must be online to generate dialogue. Offline playback is possible once you have downloaded the audio files, but generation requires connectivity.
Are the voices consistent across languages?
Yes, each voice preset maintains its tonal character across all 23 supported languages. For example, aria remains warm and clear whether speaking English, Japanese, or Spanish. This consistency helps maintain character identity even when your game supports multiple languages without needing separate voice actors.
Can I use the voices for audiobooks?
Yes, you can use the voices for audiobooks. The API supports texts up to 4096 characters per request, so you can split longer books. The generated audio is returned in the response and not stored, giving you full control over the files for your audiobook project.
Do I need to credit the voices?
The voices are based on LibriTTS-R (CC BY 4.0) and LibriVox public-domain readings. The API does not require specific attribution for the synthetic output. The generated files are yours to use in your project.
Can I clone a voice for a specific NPC?
No, we do not offer custom voice cloning. You must choose from the eight available presets: aria, nova, luna, sage, felix, milo, atlas, or orion. If you need a specific voice, you will need to select the preset that best matches your NPC’s character.
How do I handle long dialogues?
Split long texts into parts of up to 4096 characters. Generate audio for each part separately using the API, then join the resulting audio files in your game engine. This ensures reliable processing and avoids hitting the character limit for a single request.
What happens if the API goes down?
We do not provide an SLA or uptime guarantee. If the API is unavailable, you cannot generate new audio until it is back online. Failed requests are free, so you are not charged for attempts made during downtime. Plan for occasional interruptions in your workflow.
Can I use the API for live streams?
You can use the API for live streams, though generation takes time. However, generation takes time, roughly real-time for long texts. For live use, pre-generate dialogue or accept slight delays. The API processes requests sequentially, so high load may cause queuing delays during peak times.
Do I need to install software?
No, you do not need to install any software. The API is accessed via HTTP requests. You can use the official OpenAI SDKs or plain HTTP to send text and receive audio. It works with any programming language that can make web requests.
Is the API easy to use?
Yes, the API is easy to use. It follows the same request shape as the OpenAI speech endpoint. You can use existing OpenAI SDKs by changing the base URL. A ready curl example is provided on the dashboard to help you start generating dialogue quickly.
Choosing the Right Voice Tone
Selecting the right voice preset is critical for immersion in indie games. Each of our eight presets offers a distinct tonal quality, from warm narrators to energetic announcers. This allows you to match the voice directly to the narrative context without needing emotion tags or custom cloning.
The API provides straightforward presets like Aria for lore or Felix for promos. By pairing the correct voice with the right script, you create consistent character identities that resonate with players, keeping them engaged without the legal headaches of traditional voice acting.
Warm Narrators for Lore
Voices like Aria and Luna bring warmth and clarity to ancient stories or background narration. Their clear, storytelling tones help players absorb complex world-building details without feeling overwhelmed. These presets are ideal for reading extended text passages where a gentle, consistent tone maintains immersion.
Use these voices for environmental storytelling or historical context. The soft, engaging delivery keeps players focused on the narrative flow, ensuring that lore drops feel natural rather than mechanical.
- Aria: Warm, clear narrator ideal for ancient history or mystical tales.
- Luna: Soft, storytelling voice perfect for emotional or atmospheric background narration.
Energetic Promos for Trailers
For trailers or quick UI prompts, Felix delivers an energetic, promo-ready tone. His voice is designed to grab attention quickly, making him perfect for short, punchy text that needs to convey excitement or urgency. This energy helps highlight key game features or limited-time events.
Pair Felix with dynamic visual cuts to create a cohesive promotional feel. His voice stands out clearly against background music, ensuring the message is heard even in noisy gameplay environments.
- Felix: Energetic, promo voice ideal for trailers, updates, or urgent notifications.
Calm Instructors for Tutorials
Sage offers a calm, instructional tone that guides players through mechanics without sounding robotic. This preset is excellent for tutorial systems, helping new users understand controls or rules with a friendly, approachable demeanor. Milo also works well for conversational, friendly guidance.
Use Sage for step-by-step instructions where clarity is paramount. The neutral yet warm delivery reduces player frustration by making the learning process feel like a helpful companion rather than a dry manual.
- Sage: Calm, instructional voice ideal for tutorials and system explanations.
- Milo: Friendly, conversational voice suitable for interactive NPC guidance.
Voice Presets Overview
VoxQuest offers eight distinct voice presets, each designed to fit specific NPC archetypes. These voices are derived from licensed LibriTTS-R recordings, ensuring clear, usable audio for your game's dialogue system without complex licensing overhead.
Female Presets: Aria, Nova, Luna, Sage
Aria provides a warm, clear narration ideal for lore-heavy quests. Nova brings a bright, upbeat energy perfect for shopkeepers or cheerful companions. Luna offers a soft, storytelling tone for atmospheric scenes, while Sage delivers a calm, instructional voice ideal for tutorial NPCs or guides.
Male Presets: Felix, Milo, Atlas, Orion
Felix is energetic and promo-ready, great for hype-driven NPCs or announcers. Milo sounds friendly and conversational, suitable for everyday characters. Atlas provides a neutral, news-style tone for objective narrators, and Orion offers a deep, documentary voice for serious lore or historical figures.
Gender and Tone Variations
Any voice can speak any of the 23 supported languages. You can adjust speed between 0.25x and 4.0x to fine-tune pacing for different character personalities or dialogue density needs.
Each generated audio file includes an inaudible watermark, marking it as synthetic speech. This helps maintain transparency while keeping the audio clean and professional for your players.
Language Support for Global Games
Supporting a global audience means your NPCs can speak the language your players expect. Our API handles 23 languages, allowing you to localize dialogue without hiring separate voice actors for every region. This makes it easy to expand your indie game’s reach while keeping production costs predictable.
English and Spanish
Start with the most widely spoken languages. English and Spanish cover a massive portion of the global gaming market. You can generate dialogue for main quests and side conversations in these languages using any of our eight voice presets.
- Clear narration for main storylines
- Conversational tones for casual interactions
Asian Languages: Japanese, Korean
Expand into high-demand markets with Japanese and Korean support. Since the API guesses language from script, you must set the ISO code explicitly for Latin-script text to ensure correct pronunciation. This ensures your NPCs sound native, not like they are reading English with an accent.
- Native pronunciation for key markets
- Explicit language code for Latin text
European Languages
Dive deeper into regional preferences with support for German, French, Italian, and more. Whether you are building a strategy game for European players or a narrative adventure, you can mix and match voices across languages to create a rich, diverse world.
- Support for 23 total languages
- No extra cost for additional languages
Technical Integration
To generate NPC dialogue, send a POST request to https://api.playmm9.com/v1/audio/speech with your text and chosen voice. The API returns an audio file directly in the response, which you can save to your project’s assets folder.
Use the official OpenAI SDK or plain HTTP to handle the integration. Ensure your key is passed in the Authorization header, and handle the 60 requests per minute limit by batching text if necessary for large dialogue trees.
API Setup for Unity
Use Unity’s HTTPClient or WWWForm to send the JSON payload. Set the base URL to https://api.playmm9.com/v1. Parse the returned audio bytes into an AudioClip for immediate playback in your game engine.
Remember to split long dialogue into chunks of 4096 characters. Concatenate the resulting audio clips in Unity to maintain continuous speech for longer NPC responses.
- Use POST https://api.playmm9.com/v1/audio/speech
- Pass text, voice, and format in JSON body
- Receive audio bytes for AudioClip conversion
API Setup for Unreal
In Unreal Engine, use the HTTP module to send requests to the API endpoint. Authenticate with your Bearer token. The response contains the audio file data, which you can save to disk.
For Unreal, consider using the WAV format for lower latency during playback. You can also generate MP3 files for storage to save disk space, converting only when needed for real-time dialogue.
- Authenticate with Bearer token
- Send JSON payload via HTTP module
- Process audio response for UE audio components
Web Implementation
For web-based games, use fetch() to call the API. Handle the response as a Blob or ArrayBuffer. Create an audio object from the data for playback in the browser.
Ensure you handle CORS correctly if hosting your game on a different domain. The API supports standard web audio formats, making integration straightforward for browser-based projects.
- Use fetch() for API calls
- Handle response as Blob or ArrayBuffer
- Create Audio object for playback
Cost-Effective Dialogue Systems
Managing NPC dialogue volume requires balancing audio quality with character limits. Our API charges $10 per million characters, making bulk generation highly affordable for large dialogue trees. Each request supports up to 4,096 characters, allowing you to batch entire conversation branches efficiently.
For shorter phrases, generation takes a few seconds, but long texts process at roughly real-time speed. Under load, requests queue sequentially, so plan your generation pipeline accordingly. Failed requests incur no charge, ensuring you only pay for usable audio.
Bulk Text Generation
Split longer scripts into 4,096-character chunks and join the resulting audio files. This approach keeps individual requests within limits while maintaining narrative continuity for extended monologues or complex quest lines.
Since characters include spaces and punctuation, accurate counting helps predict costs precisely. With a free trial of $0.50 (approx. 50,000 characters), you can test bulk workflows before committing to a top-up.
- Supports up to 4,096 characters per request
- Charges $10 per 1 million characters
- Failed requests are free
Speed Optimization
Adjust the speed parameter (0.25–4.0) to control generation time and voice pacing. Faster speeds reduce processing time for rapid-fire NPC exchanges, while slower speeds enhance clarity for emotional or instructional dialogue.
Requests are processed sequentially, so high traffic may introduce queue delays. For real-time integration, pre-generate dialogue assets during load screens or use asynchronous calls to keep your game responsive.
- Adjustable speed from 0.25 to 4.0
- Processes requests sequentially
- Real-time speed on long texts
Format Selection
Choose mp3 for compact size, wav for uncompressed quality, opus for efficient compression, flac for lossless fidelity, or pcm for raw data. Format choice impacts storage and bandwidth without affecting voice quality.
Use mp3 or opus for mobile games to conserve storage. Reserve flac or wav for high-fidelity desktop releases where audio clarity is prioritized over file size.
- Available formats: mp3, wav, opus, flac, pcm
- MP3/OPUS for efficient storage
- FLAC/WAV for high fidelity
Licensing and Rights
VoxQuest’s synthetic voices are derived from licensed recordings under the LibriTTS-R license (CC BY 4.0) and sourced from LibriVox public-domain audiobooks. This foundation ensures that your NPC dialogue carries a clear, traceable lineage without hidden royalty clauses, allowing you to integrate generated audio directly into commercial game builds without fear of copyright strikes.
LibriTTS-R License
The underlying voice models are trained on the LibriTTS-R dataset, which operates under the Creative Commons Attribution 4.0 International license. This means the core voice data is legally shareable and modifiable, providing a stable foundation for your game’s audio assets.
- Based on CC BY 4.0 licensed data
- Transparent data lineage for all voices
Public Domain Audio
Many source recordings originate from LibriVox, a project dedicated to public-domain audiobooks. Because these source performances are in the public domain, they do not carry restrictive copyright encumbrances that could complicate distribution on platforms like Steam or itch.io, giving you peace of mind when shipping your indie title.
Commercial Use Rights
You retain full rights to use the generated audio files in your games, whether free or paid. There are no per-seat fees or revenue caps tied to the voice license itself. The only requirement is that you use the generated files as part of your final product, with no additional legal overhead for NPC dialogue or narration.
- No per-seat or revenue caps
- Safe for Steam, itch.io, and console releases
Audio Quality and Formats
Selecting the right audio format balances file size, quality, and platform compatibility. Our API lets you choose between MP3, WAV, OPUS, and FLAC, ensuring your NPC dialogue fits your game’s technical constraints without compromising on clarity.
Each format serves a specific purpose, from lightweight web delivery to high-fidelity archival. Understanding these options helps you optimize load times while maintaining immersive voice acting standards for your players.
MP3 vs WAV
MP3 is the standard for most games due to its small file size and broad compatibility. It offers good quality at low bitrates, making it ideal for dialogue with many lines. WAV provides uncompressed, lossless audio, perfect for high-quality cutscenes where file size is less critical than pristine sound.
- MP3: Suitable for general NPC dialogue and limited storage.
- WAV: Suitable for premium cutscenes and high-fidelity moments.
OPUS for Web
OPUS is an excellent choice for web-based games or browser ports. It delivers high-quality audio at very low bitrates, reducing bandwidth usage significantly. This makes it perfect for maintaining performance on lower-end devices or unstable connections.
- Optimized for low bandwidth scenarios.
- High quality even at small file sizes.
FLAC for High Fidelity
FLAC (Free Lossless Audio Codec) offers lossless compression, preserving every detail of the generated voice. Use this format for premium audio packages or when you need to archive master files without quality loss. It is smaller than WAV but retains full fidelity.
- Lossless compression for perfect audio reproduction.
- Ideal for master archives and premium tiers.

Rate Limits and Scaling
Your API key allows 60 requests per minute, scaling up to 10 requests per minute before your first top-up. This limit applies per key, so managing concurrent synthesis jobs for NPC dialogues requires careful planning to avoid throttling during peak gameplay moments.
Handling High Traffic
When generating bulk dialogue for a large cast, batch your requests. Since requests are processed sequentially, high concurrency may cause queuing. Distribute synthesis tasks over time or use multiple keys if your account supports it to maintain steady throughput.
- Monitor your current requests per minute via the dashboard.
- Pre-generate static NPC lines during load screens.
- Avoid sending all dialogue requests simultaneously at game start.
Queue Management
Requests are processed one after another, so under load a request may wait in a queue. For short phrases, generation takes a few seconds; for longer texts, expect real-time speeds (approx. 25 characters per second). Plan your audio assets accordingly.
- Prioritize critical voice lines over ambient chatter.
- Use asynchronous API calls to keep the game responsive.
- Implement local fallback audio if the queue exceeds 10 seconds.
Retry Logic
If a synthesis fails, the charge is returned in full. Implement exponential backoff for retries. Always check for HTTP 429 Too Many Requests errors and pause briefly before resuming your speech generation loop.
- Handle 429 errors with a 1-second delay.
- Log failed requests for debugging.
- Verify audio integrity before playing back generated clips.
Common Use Cases
VoxQuest’s text-to-speech API transforms static text into dynamic audio, allowing indie developers to populate worlds without recording dozens of voice actors. By leveraging eight distinct voice presets, you can create consistent characters for dialogue trees, ensuring every line sounds unique yet cohesive across your entire game.
This approach is particularly valuable for visual novels and procedural quests, where volume of dialogue can strain small budgets. With a $10 per million characters rate and a free trial credit, you can generate extensive NPC lines affordably, keeping your project’s scope manageable while maintaining high production value.
RPG Dialogue Trees
Use the API to generate contextual responses based on player choices. With voices like Sage for calm instructions or Felix for energetic promos, you can differentiate quest givers from shopkeepers easily. The 4,096 character limit per request fits most standard dialogue lines, and you can split longer conversations into manageable chunks.
Since the API returns audio files directly, you can integrate them into your game engine via simple HTTP calls. This allows for dynamic speech where NPCs react to specific player actions, enhancing immersion without the need for pre-recorded voice banks.
- Assign unique voices to major factions or classes
- Generate dynamic responses based on player stats
- Use Sage for tutorial prompts and Atlas for news updates
Visual Novel Narration
Visual novels often require extensive narration that can be costly to record. VoxQuest’s Luna offers a soft, storytelling tone perfect for internal monologues, while Aria provides a warm narrative voice for exposition. You can generate entire chapters’ worth of dialogue for a fraction of traditional voice acting costs.
Since input text is not used for training, your creative work remains secure. The inaudible watermark marks the audio as synthetic, which is transparent for distribution. You can mix these generated tracks with your music and sound effects seamlessly.
- Use Luna for internal thoughts and Aria for exposition
- Generate large volumes of dialogue quickly
- Maintain consistent character voices across chapters
Procedural Quests
For procedural games, every quest might be unique. The API allows you to generate voice lines on the fly, ensuring no two quests sound identical. With support for 23 languages, you can localize quests automatically without hiring translators and voice actors for each language.
Use the speed parameter to adjust the pace of delivery, making NPCs sound excited or rushed depending on the quest context. This flexibility ensures that even randomly generated dialogue feels intentional and polished, enhancing the player’s experience.
- Generate unique voice lines for every quest
- Localize quests into 23 languages instantly
- Adjust speech speed to match quest urgency
Troubleshooting
Game developers often encounter predictable hurdles when integrating synthetic NPC voices. Understanding the common API responses and audio artifacts helps you build resilient dialogue systems that handle edge cases gracefully without disrupting the player experience.
Common API Errors
Most errors stem from exceeding the 4096 character limit or hitting the 60 requests per minute rate limit. If you receive a 429 error, implement exponential backoff in your retry logic to queue subsequent requests without overwhelming the endpoint.
- 400: Invalid JSON or missing required fields like 'model' or 'voice'.
- 401: Expired or incorrect Bearer token; regenerate via /get-api-key/.
- 429: Rate limit exceeded; wait and retry with backoff.
- 500: Internal synthesis error; retry once, as failed requests are free.
Audio Glitches
Synthetic speech may occasionally produce artifacts if the input text contains unusual punctuation or non-standard characters. Ensure your text is cleaned before sending, and test edge cases with different response formats like WAV or FLAC to identify compression artifacts.
- Stuttering: Usually caused by rapid successive requests; add small delays between calls.
- Silence: Check if input text is empty or exceeds 4096 characters.
- Noise: Try switching from MP3 to WAV for debugging quality issues.
Language Detection Issues
If the 'language' field is omitted, the API guesses the language based on the script. Latin script defaults to English, which causes mispronunciation for other languages like Spanish or German. Always explicitly set the 'language' ISO code for non-English text to ensure accurate pronunciation.
- Explicitly set 'language' for Latin-script non-English text (e.g., 'es', 'de').
- Verify supported languages: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish.
- Unsupported languages include Vietnamese, Indonesian, Thai, Ukrainian, and Romanian.
Future-Proofing Your Game
Design your audio pipeline to handle growing dialogue trees without manual re-imports. Store generated audio in object references or external assets, allowing the API to regenerate assets if voice presets change or if you need to refresh narration for updated quest text.
Scalable Architecture
Implement a modular audio manager that decouples text strings from audio files. This allows you to swap voice providers or update presets without breaking your game’s dialogue system. Use asynchronous loading to prevent frame drops during heavy dialogue sequences.
- Use object references for audio clips to simplify updates.
- Implement async loading to maintain performance during dialogue bursts.
Updating Voices
Since voices are presets based on licensed recordings, you can switch between Aria, Nova, or Milo across your entire project by updating the voice parameter in your API calls. Re-generate affected audio files to reflect new tone choices without rewriting dialogue scripts.
- Change the
voicefield in API requests to update tone globally. - Regenerate only affected assets to save time and credits.
Expanding Languages
Support global releases by leveraging the 23 supported languages. Since any voice can speak any language, you can maintain consistent character personalities across different regions by simply passing the language ISO code alongside your text input.
- Pass the
languageISO code to localize dialogue instantly. - Maintain character consistency across regions using the same voice preset.
Conclusion
VoxQuest delivers reliable, legally safe NPC dialogue for indie developers. Our API integrates seamlessly into Unity and Unreal, offering clear presets and transparent pricing without hidden fees.
Summary of Benefits
Generate synthetic speech. Access 8 distinct voices across 23 languages. Pay only for what you use, with no monthly commitments or minimum spend.
- Affordable at $0.01 per 1,000 characters
- Instant API integration
- Clear commercial rights
Next Steps
Sign up on /get-api-key/ to receive your $0.50 trial credit. Test the API with our curl example or integrate the OpenAI SDK. Top up via USTRC-20 for ongoing production.
Final Thoughts
Build immersive worlds without voice acting headaches. VoxQuest simplifies synthetic speech for your next game project. Start generating dialogue today.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. This preset is excellent for tutorial systems, helping new users understand controls or rules with a friendly, approachable demeanor.
23 languages
Any voice speaks any of these languages.
