About Linguicore: Synthetic Speech API for ML Teams

By Linguicore editorial team · Updated 2026-11-18

This page explains who Linguicore is and why we built a dedicated text-to-speech API for teams creating synthetic speech datasets for machine learning. We provide consistent, watermark-marked audio in 23 languages, designed to streamline your data generation workflow without the complexity of managing large infrastructure.

As an independent provider, we focus on delivering reliable, transparent API access for developers who need volume and consistency.

About Linguicore: Synthetic Speech API for ML Teams

Frequently asked questions

Are you affiliated with OpenAI?

No. We are an independent storefront for a text-to-speech API. Our endpoint structure mimics the OpenAI speech endpoint for compatibility, but we are not affiliated with OpenAI, ElevenLabs, Google, Amazon, Microsoft, or any other brand mentioned in comparisons.

Who owns the voices?

Our voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and public-domain audiobook readings from LibriVox. We use these recordings as the basis for our eight preset voices, which are then processed through our TTS engine to generate audio files.

Do I need to attribute the voices?

The source material is licensed under CC BY 4.0, which typically requires attribution. We do not provide custom voice cloning, so you are using our preset voices.

Can I resell the audio?

Yes. You can use the generated audio for your machine learning datasets or other commercial purposes. The audio is returned in the response and not stored by us. Each file carries an inaudible watermark marking it as synthetic speech,

How do you handle data privacy?

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 (time, characters, status) solely for billing purposes. We do not offer GDPR certificates.

Do you have a physical address?

We do not list a physical address. We are a cloud-based API service. We do not have phone support or live chat. All communication and support are handled via email. You can find contact details on our /contact/ page.

Who runs the editorial team?

Linguicore is an independent storefront. We do not have a large editorial team. Our content is managed directly by the team behind the API. We focus on providing honest, specific documentation and updates without hype or superlatives about our synthetic speech capabilities.

Can I request a new voice?

We currently offer eight preset voices (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion). We do not offer custom voice cloning or voice design. You cannot upload your own voice to create a custom preset. The list of voices is fixed.

Do you use AI to generate voices?

Our voices are based on licensed recordings from LibriVox and processed through our text-to-speech engine. The voices are derived from existing public-domain and licensed audio sources.

How do I contact support?

Support is by email only. We do not have a phone number or live chat. You can reach us via the contact form on our /contact/ page. We aim to respond to your inquiries regarding your API usage or billing questions promptly.

Is the company scalable?

Our API handles requests sequentially, so under load, a request may wait in a queue. We do not offer an SLA or uptime guarantee. However, we are designed to handle the needs of large teams building synthetic speech datasets without complex infrastructure.

Do you have investors?

We are an independent storefront. We do not disclose investor information. Our business model relies on direct API usage fees and top-ups via USDT on the TRON network. We do not have monthly plans, seats, or minimum spend requirements.

What is your mission?

We provide a simple text-to-speech API designed for teams building synthetic speech datasets for machine learning. Our goal is to help you generate consistent, watermark-marked audio files in 23 languages to train your models without complex infrastructure, focusing on reliability and simplicity.

Do you offer refunds?

Failed requests are free; if synthesis fails, the charge is returned in full. We do not offer refunds for successful requests. Balances do not expire. You can top up with a minimum of $5 via USDT on the TRON network.

Who We Are

Sound wave illustration for: Who We Are

We are an independent team building a text-to-speech API specifically for machine learning teams. We focus on delivering consistent synthetic speech datasets without the complexity of enterprise infrastructure.

Independent API provider

Independent API provider

We operate as a dedicated service, not a division of a larger tech giant. Our infrastructure is built solely for generating audio files from text, ensuring reliability and direct access to our core capabilities.

Focused on ML teams

Focused on ML teams

Our API is designed for teams that need to train models with large volumes of synthetic data. We provide 23 languages and consistent voice presets, allowing you to generate diverse training sets efficiently.

  • Optimized for bulk generation workflows
  • Consistent voice characteristics across files
  • Standard JSON API compatible with existing tools

No corporate overhead

No corporate overhead

Without the layers of a massive corporation, we keep our pricing simple and transparent. You pay only for what you use, with no hidden fees or complex tiered plans.

Our structure allows us to respond directly to developer needs without bureaucratic delays.

Voice Licensing

Sound wave illustration for: Voice Licensing

We built our voice library on licensed recordings from LibriTTS-R (CC BY 4.0) and public-domain audiobook readings from LibriVox. This foundation allows us to offer eight distinct voice presets.

LibriTTS-R (CC BY 4.0)

LibriTTS-R (CC BY 4.0)

Our voices are derived from the LibriTTS-R dataset, which uses a Creative Commons Attribution 4.0 International license. This open license structure applies to the source recordings.

LibriVox public domain

Many of our voice models are based on public-domain readings from LibriVox. Because these source recordings are in the public domain, they provide a stable foundation for our TTS engine.

Inaudible watermarks

Every generated audio file contains an inaudible watermark that identifies the speech as synthetic. This feature is crucial for ML teams training models on mixed datasets, as it allows you to accurately label synthetic training data and maintain transparency in your data pipelines.

Data Policy

Sound wave illustration for: Data Policy

Linguicore provides a text-to-speech API designed for teams building synthetic speech datasets for machine learning. Generate consistent, watermark-marked audio files in 23 languages to train your models without complex infrastructure.

Input text not used for training

Your input text is used exclusively to produce the audio file. We do not use your source text for training our models or improving the engine. Your data remains yours, ensuring privacy for your proprietary datasets.

Audio not stored

Generated audio is returned directly in the API response and is not stored on our servers after delivery. This minimizes data retention risks and keeps your synthetic speech assets strictly under your control.

Metadata kept for billing

We retain request metadata, including timestamp, character count, and status, solely for billing purposes. This ensures accurate charging based on the $0.01 per 1,000 characters rate and helps track usage limits.

Technical Limits

Sound wave illustration for: Technical Limits

Our API is designed for batch processing synthetic speech datasets, so it does not support streaming audio. You send a request and receive the complete file once synthesis finishes. For long texts, you split them into 4,096-character chunks and join the resulting audio files manually. Requests are processed sequentially, so high volume may cause queueing delays.

We do not support SSML, emotion tags, or custom pronunciation lexicons. The engine relies on standard text input and optional language codes. If you need precise phonetic control, you must manage it in your input text before sending it to the API.

No streaming support

The API returns the final audio file after processing is complete. It does not stream partial audio chunks in real-time. This design simplifies integration for dataset generation but means you cannot play audio while it is being generated.

For texts exceeding 4,096 characters, you must split the input, generate separate audio files, and concatenate them on your end. Under heavy load, requests are queued, so generation time may exceed real-time speed (roughly 25 characters per second).

No SSML

We do not support Speech Synthesis Markup Language (SSML). You cannot use tags to control pauses, emphasis, or pronunciation directly in the request body. The engine interprets standard text and guesses language from the script if not specified.

No voice cloning

We offer eight fixed voice presets based on licensed recordings. We do not provide custom voice cloning, so you cannot upload a reference sample to create a unique voice. The voices are synthetic interpretations, not exact replicas of specific individuals.

Technical Limits

Support Contact

Sound wave illustration for: Support Contact

We keep support simple because our API is straightforward. Most issues are resolved by checking the documentation or adjusting request parameters like speed or language codes.

We believe in letting the product speak for itself through reliability and clear error messages, rather than complex support tiers.

Email support only

All inquiries go to a single email inbox. We do not use ticketing systems that add layers of bureaucracy.

You can reach us at [email protected] for billing questions, technical issues, or general inquiries.

No phone or chat

There is no phone number and no live chat widget on this site. We avoid real-time interruptions to keep our infrastructure focused on processing your speech requests efficiently.

This also means you won't get put on hold while waiting for an agent to become available.

Response times

We aim to reply to emails within 24 hours during business days.

Since we are a small, independent team, responses are direct and technical, avoiding generic customer service scripts.

Why We Exist

Sound wave illustration for: Why We Exist

We built Linguicore because generating consistent synthetic speech for ML datasets shouldn’t require complex infrastructure. Teams need reliable, watermark-marked audio in 23 languages to train models without the overhead of managing voice licenses or synthesis servers.

Democratizing synthetic data

Our voices are licensed from LibriTTS-R (CC BY 4.0) and LibriVox public-domain recordings. This lets you generate large-scale datasets without legal friction. Every file carries an inaudible watermark, clearly marking synthetic speech for your training pipelines.

Simplifying API access

We provide a simple POST endpoint that mirrors the OpenAI format. You can use standard OpenAI SDKs by just changing the base URL. No custom SDKs or complex integrations are needed to start generating audio immediately.

Transparent pricing

Cost is strictly $10 per 1 million characters. There are no monthly plans, seats, or hidden fees. You pay only for what you generate, and failed requests are free. Top-ups via USDT on TRON are processed automatically.

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. This ensures data provenance and prevents accidental mixing of human and synthetic samples during model training.

Avatar of the voice aria
ariafemale
Avatar of the voice nova
novafemale
Avatar of the voice luna
lunafemale
Avatar of the voice sage
sagefemale
Avatar of the voice felix
felixmale
Avatar of the voice milo
milomale
Avatar of the voice atlas
atlasmale
Avatar of the voice orion
orionmale

23 languages

Any voice speaks any of these languages.

Sources and further reading

  1. Wikipedia
  2. openai-node on GitHub
  3. FFmpeg documentation
  4. curl documentation
  5. MDN: Web Speech API
  6. W3C WAI: making audio and video accessible
  7. OpenAI API reference: create speech
  8. Node.js documentation
  9. LibriTTS-R paper (arXiv 2305.18802)
  10. LibriVox public-domain audiobooks
  11. Python documentation
  12. ISO 639 language codes

More on this site