Frequently asked questions
Who owns the API?
NPC Audio is an independent storefront for the text-to-speech API. We are not affiliated with OpenAI, ElevenLabs, or any other major brand. Our infrastructure and billing are managed directly by NPC Audio for game developers who need consistent, licensable NPC dialogue without licensing overhead.
Where are the voices recorded?
Our voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and public-domain audiobook readings from LibriVox. We provide eight presets derived from these sources, ensuring clear, licensable audio for your game’s dialogue without requiring additional voice actor contracts.
Is the team based in the US?
The provided facts do not specify the geographic location of the NPC Audio team. We operate as an independent storefront for the API, focusing on engineering and concise delivery for game developers rather than regional office presence or local support teams.
Do you store my game data?
Input text is used only to produce the audio and is not used for training. Generated audio is returned in the response and not stored. We keep request metadata, such as time, characters, and status, strictly for billing purposes.
How do I contact support?
Support is available by email only. We do not offer a phone number, live chat, or on-call engineering team. You can reach us via the contact page. Responses are handled asynchronously, so please include your API key and details in your message.
Can I request a new voice?
We currently offer eight fixed presets: Aria, Nova, Luna, Sage, Felix, Milo, Atlas, and Orion. We do not offer custom voice cloning or voice design services. You cannot upload your own voice or request new presets; you must choose from the existing library.
Do you have a privacy policy?
We do not explicitly list a separate privacy policy document in our facts. However, we state that input text is not used for training and generated audio is not stored. Request metadata is retained solely for billing. We do not claim specific certifications like GDPR or SOC 2.
Are the voices royalty-free?
The voices are based on LibriTTS-R (CC BY 4.0) and LibriVox public-domain readings. We provide the synthesis engine; you handle the narrative design and integration into your game.
Do you offer refunds?
Failed requests are free: if synthesis fails, the charge is returned in full. We do not offer general refunds for successful generations. Payments are made via USDT on the TRON network. Balance does not expire, and there are no monthly plans or subscriptions.
Is the API open source?
We do not offer an official SDK of our own. You use the standard OpenAI SDK or plain HTTP by changing the base_url. We provide documentation and curl examples, but the underlying synthesis engine and infrastructure are proprietary.
Can I use the API commercially?
Yes, the API is designed for game developers. You can use the synthesized audio for NPC dialogue in commercial games. We handle the synthesis engine so you can focus on narrative design. There are no minimum spend requirements or enterprise contracts limiting your commercial use.
Do you have a status page?
We do not have a dedicated status page. We do not offer an SLA or uptime guarantee. Requests are processed sequentially, so under load, a request may wait in a queue. You can monitor your own request status via the API response codes.
How do I report a bug?
You can report bugs via email support. Please include your API key, the request payload, and the response you received. Our team reviews emails and addresses issues asynchronously without a dedicated engineering queue.
Do you accept GitHub issues?
No, we do not accept GitHub issues. Support is handled exclusively via email. We do not maintain a public issue tracker for the API. Please send bug reports or feature requests to our support email address with your account details and reproduction steps.
Who we are
NPC Audio is an independent text-to-speech API provider built for game developers who need consistent, licensable NPC dialogue without licensing overhead. We handle the synthesis engine so you can focus on narrative design and integration.
Our product is a standard cloud API. You send text, receive an audio file. We do not offer a proprietary SDK or mobile library; you use the standard OpenAI SDK or plain HTTP.
Independent API provider
We operate as a standalone storefront for the same TTS engine used across our network. This means you get direct access to the API without intermediaries or reseller markup layers.
Focus on game developers
We designed the voice presets and pricing structure specifically for game dialogue workflows. Our 8 voice presets cover the common needs of NPC interactions, from narration to conversation.
- 8 presets tailored for dialogue
- Simple per-character pricing
- Standard HTTP integration
No corporate overhead
We are not a large enterprise. This keeps our operations simple and our pricing transparent. You pay only for what you use, with no minimum spend or monthly plans.
Voice licensing
NPC Audio’s voice engine is built on licensed recordings from the LibriTTS-R dataset. These voices are based on licensed recordings from the LibriTTS-R dataset (CC BY 4.0), which uses data from LibriVox public-domain audiobook readings.
LibriTTS-R source
The voice presets are based on licensed recordings from the LibriTTS-R dataset. The dataset provides the source recordings for our eight voice presets.
CC BY 4.0 license
LibriVox releases its recordings under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. The source recordings are licensed under CC BY 4.0.
Public domain readings
The underlying readings are in the public domain, meaning the original performances are free of copyright restrictions.

Data policy
Input text is used exclusively to generate the requested audio file and is not retained after synthesis. The generated audio is returned in the response and is not stored on our servers, ensuring that your dialogue assets remain transient unless you save them.
Since input text is not stored, it is not used for training. The API treats each request as an isolated transaction, preserving the confidentiality of your game’s narrative content without contributing it to any model improvement datasets.
Input text not stored
Once the audio file is generated and returned to your request, the source text is discarded. We do not keep a copy of the text you send, ensuring that your dialogue scripts remain private and are not archived for future use.
No training data usage
The text you submit for synthesis is never added to our training data. Unlike some consumer-facing services, NPC Audio does not use your input to improve the model, keeping your content strictly separate from our engine’s development.
Metadata kept for billing
We retain minimal request metadata—including timestamp, character count, and status—solely to calculate charges and verify usage limits. This data is used exclusively for billing accuracy and does not include the content of your dialogue or audio files.
Team and contact
NPC Audio operates as a lean, independent engineering team focused exclusively on delivering reliable text-to-speech infrastructure for game developers. We prioritize technical stability and clear documentation over corporate bureaucracy, ensuring that our API remains stable and predictable for your NPC dialogue pipelines.
Our structure is designed to minimize overhead and keep costs low. By maintaining a small, specialized team, we can respond directly to technical questions and ensure that our synthesis engine remains optimized for the specific needs of game audio integration.
Editorial team
They ensure that voice descriptions accurately reflect the synthesis engine's capabilities and that our technical guides remain clear and up-to-date for integration workflows.
Email support only
We provide support via email to maintain detailed records of your technical issues. This approach allows our engineers to review your specific integration context and provide precise, documented solutions without the noise of live chat environments.
No live chat
We do not offer live chat support. This decision helps us keep operational costs low and ensures that you receive thorough, written responses that can be easily referenced during your development cycle.

Technical approach
NPC Audio uses standard HTTP requests, making integration straightforward. You can send JSON bodies via POST to the speech endpoint and receive audio files directly. This approach works with any HTTP client, from simple curl commands to complex game engine scripts.
OpenAI SDK compatible
Our API mirrors the OpenAI speech endpoint structure. By updating the base_url to https://api.playgperya.com/v1 and using your NPC Audio key, you can leverage official OpenAI SDKs without modification. This reduces boilerplate code for common languages like Python, JavaScript, and C#.
- Use standard OpenAI SDKs with a simple base_url change
- Compatible with Python, JavaScript, C#, and other language clients
Standard HTTP API
The API accepts JSON payloads with fields like model, input, voice, and response_format. Responses return audio files in formats such as mp3, wav, opus, flac, or pcm. You handle the text input, and we return the synthesized audio file directly in the response stream.
- JSON input with standard fields
- Multiple audio format outputs (mp3, wav, opus, flac, pcm)
No official SDK
We do not maintain a proprietary NPC Audio SDK. Instead, we rely on the robust, well-documented OpenAI SDKs and standard HTTP libraries. This means you get community support and familiar tooling, without needing to learn a new library specific to our platform.

Product limitations
NPC Audio is a focused text-to-speech API designed for game developers who need consistent, licensable NPC dialogue without licensing overhead. We handle the synthesis engine so you can focus on narrative design and integration.
Our approach relies on standard HTTP requests and the OpenAI SDK, ensuring straightforward integration without proprietary dependencies. This simplicity reduces friction during the development cycle, allowing your team to prioritize narrative and gameplay mechanics.
No streaming audio
Requests are processed sequentially, meaning you receive a complete audio file rather than real-time partial streams. For longer texts, split the input into parts of up to 4096 characters and join the resulting audio files manually. This ensures deterministic output suitable for game dialogue systems.
- Standard HTTP POST response
- Requires manual joining for long texts
- No WebSocket or partial audio chunks
No SSML support
The API accepts plain text input without SSML markup for precise control over pauses, emphasis, or pronunciation. Speed adjustments are handled via the speed parameter (0.25–4.0) rather than granular prosody tags. This keeps the request body simple and reduces parsing overhead.
No custom voice cloning
We do not offer custom voice cloning for customers, nor do you upload your own voice recordings. The API uses eight preset voices derived from licensed LibriTTS-R recordings under CC BY 4.0. These voices are consistent and licensable, eliminating the need for complex voice management pipelines.
- Eight fixed presets only
- No user-uploaded voice samples
- No custom voice design
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. The game developer hired a freelancer to voice the quirky NPCs in their new fantasy title.
23 languages
Any voice speaks any of these languages.
