Frequently asked questions
Who runs LinguaForge?
LinguaForge is an independent storefront for the text-to-speech API hosted at 2tiger.info. We operate as a standalone entity providing cloud-based synthesis services. We are not affiliated with OpenAI, ElevenLabs, or any other major technology brand. Our infrastructure supports indie developers and machine learning pipelines directly through our API endpoints.
Are the voices copyrighted?
The voices are based on licensed recordings from LibriTTS-R and LibriVox public-domain audiobook readings. While the underlying source material is public domain, the specific synthetic audio files generated by our API carry an inaudible watermark marking them as synthetic speech. This ensures clear identification of the synthetic nature of the content.
Do you use my text for training?
No. Input text is used exclusively to produce the requested audio file. We do not use your input text for training our models or improving our services. Once the audio is generated and returned in the response, the input text is not retained for any training purposes.
Do you store my generated audio?
No. Generated audio is returned directly in the API response and is not stored on our servers. We retain only request metadata, such as timestamp, character count, and status, for billing purposes. The actual audio content you receive is not kept in our long-term storage systems.
How are voices licensed?
Our eight voice presets are derived from LibriTTS-R, which uses CC BY 4.0 licensed recordings from LibriVox public-domain audiobook readings. This foundation allows us to provide consistent, synthetic voices optimized for machine learning training pipelines while maintaining clear licensing origins for the underlying vocal characteristics used in our engine.
What is the CC BY 4.0 license?
CC BY 4.0 is a Creative Commons Attribution license that allows sharing and adapting the material. Our voices are based on LibriTTS-R, which uses this license for its source recordings.
Do you have an office?
We do not list a physical office address. As a cloud API provider, our operations are distributed. We do not offer on-premise installations or physical locations for client visits. All interactions, support, and service delivery occur digitally through our API endpoints and web-based account management portal.
How do I contact support?
Support is available via email only. We do not offer live chat, phone support, or ticketing systems with guaranteed response times. You can reach our team through the contact page or the email address provided in your account dashboard. Please include your API key for faster identification.
Is there a phone number?
No. We do not provide a phone number for customer support. All communication is conducted via email. This allows us to maintain a lean operational structure focused on API throughput and reliability rather than voice-based support channels. You can find our email contact details on the /contact/ page.
Do you offer enterprise contracts?
No. We do not offer enterprise contracts, SLAs, or uptime guarantees. Our service is designed for indie developers and scalable API usage without custom agreements. Pricing is standardized at $10 per million characters, with no minimum spend or volume discounts. All terms are defined in our API documentation.
Are you affiliated with OpenAI?
No. We are not affiliated with OpenAI, despite our API endpoint structure resembling theirs. We use the same request shape for compatibility, but our engine, voices, and pricing are entirely separate. We are an independent provider of synthetic speech services, distinct from OpenAI’s TTS models and infrastructure.
Do you have certifications?
No. We do not hold SOC 2, HIPAA, or GDPR certifications. We do not provide data-center choices or compliance certificates. Our data handling is straightforward: input text is used only for synthesis, and metadata is kept for billing. We do not market ourselves as a certified enterprise solution.
Where are your servers?
We do not specify the physical location of our servers. We do not offer on-premise installations or data-center selection options. Our infrastructure is fully cloud-based and managed centrally to ensure consistent API performance. You can monitor service status and usage directly through your account dashboard at /get-api-key/.
How do I report a bug?
You can report bugs via email support. Since we do not have a public issue tracker or live chat, please send details including your API key, request ID, and reproduction steps to our support email. We review all incoming reports and address issues as part of our standard maintenance cycle.
Editorial Team Overview
LinguaForge operates as an independent entity managed by a small, dedicated team focused exclusively on high-throughput synthetic speech generation. We prioritize technical precision and API reliability over expansive marketing, ensuring our infrastructure remains optimized for machine learning dataset production.
Independent Management
Our editorial and technical decisions are driven by direct engineering feedback rather than corporate hierarchy. This structure allows for rapid iteration of our TTS endpoints and consistent voice delivery, which is critical for automated labeling pipelines.
Technical Focus
We concentrate on delivering consistent, watermark-marked audio files that meet strict character-per-second generation rates. Our team maintains the underlying API infrastructure to ensure stable throughput for indie developers generating large-scale corpora without external dependencies.
No Corporate Backing
LinguaForge is not affiliated with major technology firms or cloud providers. We operate as a standalone API service, meaning our roadmap and pricing are determined solely by our operational costs and developer needs, not by shareholder mandates.
Voice Licensing Details
LinguaForge generates synthetic voices from licensed recordings of the LibriTTS-R dataset, which comprises public-domain audiobook readings sourced from LibriVox. This foundation ensures that our eight voice presets are derived from licensed recordings.
LibriTTS-R Source
The core audio data originates from the LibriTTS-R dataset, a collection of English audiobooks read by volunteers. These recordings provide the raw vocal characteristics for our eight distinct voice presets, ensuring consistent, watermark-marked synthetic speech suitable for ML training pipelines.
CC BY 4.0 License
The voices are based on licensed recordings: LibriTTS-R (CC BY 4.0), from LibriVox public-domain audiobook readings.
Public Domain Audio
By combining public domain source material with the CC BY 4.0 licensed dataset, we maintain a clear, auditable chain of title for all generated synthetic voices.
Data Usage and Privacy
LinguaForge processes input text exclusively to generate the requested audio file. The raw text is not repurposed for other services or stored indefinitely after synthesis.
Generated audio is returned directly in the HTTP response and is not retained on our servers. This ensures minimal data footprint for high-throughput corpus generation.
Input Text Utilization
Text submitted via the API is used solely for the immediate conversion into speech. We do not index or archive your source strings for secondary analysis.
No Training Data Usage
Your input text is not used to train or improve our underlying models. The generated audio files are distinct from the training data used to create the voice presets.
Audio Retention Policy
Once the audio file is delivered to your client, LinguaForge does not store a copy on our servers. Only request metadata (time, characters, status) is kept for billing purposes.
- Audio is ephemeral; no long-term storage of generated clips.
- Metadata retained strictly for billing and usage tracking.
- Input text is not archived after synthesis.
Billing and Metadata Storage
LinguaForge maintains a strict, transparent billing structure where charges are applied per character processed. Failed requests incur no cost, ensuring you only pay for successful synthesis. This model supports high-volume corpus generation without unexpected fees.
Request Metadata Retention
We retain request metadata including timestamp, character count, and status for billing reconciliation. This data is sufficient to verify usage but does not include the audio content itself.
- Timestamp of request
- Character count processed
- Request status (success/failure)
Billing Records
Billing records are tied directly to your API key and TRC-20 transaction history. You can view your balance and usage details on the /get-api-key/ dashboard at any time.
No Long-Term Audio Storage
Generated audio files are returned in the response and are not stored on our servers. Once the transfer is complete, the audio data is discarded, ensuring no long-term retention of your specific corpus samples.

Technical Limitations Disclosure
LinguaForge delivers static audio files via standard HTTP responses, ensuring reliable integration for batch processing pipelines. The API does not support real-time streaming, WebSocket connections, or progressive audio delivery, which may impact low-latency interactive applications.
No Streaming Support
Generated audio is returned as a complete file upon synthesis completion. Clients must handle the full download before playback can begin, which aligns with typical corpus generation workflows but requires buffer management in real-time scenarios.
No SSML Implementation
The API accepts plain text input without Support for Speech Synthesis Markup Language (SSML). Users cannot specify pause durations, emphasis, or pronunciation rules via markup tags, relying instead on voice presets for tonal consistency.
No Custom Cloning
LinguaForge utilizes eight fixed voice presets derived from licensed recordings. The service does not offer custom voice cloning or uploading your own voice.
- Fixed voice presets only
- No user-uploaded voice samples for cloning
- Standardized synthetic output
Contact and Support Channels
LinguaForge operates with a lean technical team, meaning support is handled directly by the engineers who maintain the API infrastructure. This structure ensures that inquiries regarding voice synthesis, billing, or metadata retention are addressed by personnel with direct knowledge of the system's architecture.
We prioritize written communication to maintain a clear, searchable record of your support tickets. This approach is standard for our indie developer audience, allowing for precise reference to API keys, request IDs, and usage logs without the overhead of voice calls.
Email Only Support
All support requests are managed exclusively via email. You can reach our team at [email protected]. Support is by email only.
Using email allows us to attach relevant logs, explain rate limit behaviors, or clarify voice characteristics without the noise of voice transmission. It also ensures that every interaction is timestamped and archived for your records.
- Address: [email protected]
- Method: Email only
- Response time: Standard business hours
No Phone Number
We do not maintain a dedicated phone line for customer inquiries. This eliminates the need for call center overhead and keeps our focus on API reliability and documentation accuracy. If you prefer asynchronous communication, email remains the most efficient channel.
For urgent issues regarding API key invalidation or billing discrepancies, email is still the primary method. Our team can provide immediate status updates via email, which is often faster than waiting on a hold line.
No Live Chat
LinguaForge does not offer live chat widgets on our website. This decision reduces latency on our main storefronts and ensures that support staff are not interrupted during active API maintenance or debugging sessions. It also keeps our operational costs low for indie developers.
You can submit detailed descriptions of your issue, including code snippets or error logs, via email. This allows for a more thorough review than a typical chat session, resulting in more accurate resolutions for complex TTS pipeline issues.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Balance key issuance with throughput needs, as each key has a fixed rate limit.
23 languages
Any voice speaks any of these languages.
