Frequently asked questions
What is the cost per character?
The cost is $0.01 per 1,000 characters, which equals $10 per million characters. This rate applies to all models, including tts-1 and tts-1-hd, as they use the same engine and pricing structure. There are no volume discounts or tiered pricing levels; the rate remains constant regardless of usage volume.
Do spaces count toward the character limit?
Yes, spaces and punctuation count as characters. The character limit is 4,096 characters per request, including all whitespace and special characters. Ensure your input text, including spaces, does not exceed this limit to avoid request failures or truncation during synthesis.
Is there a minimum top-up amount?
The minimum top-up amount is $5. Payments are processed via USDT on the TRON network (TRC-20). Transfers below this threshold may not be processed. The minimum ensures that transaction fees do not disproportionately affect your account balance upon credit addition.
How long does USDT transfer take?
Credits are credited automatically within minutes after the transfer is confirmed on the blockchain. USDT on the TRON network typically processes quickly. Once the transaction is verified, your balance updates immediately without manual intervention or waiting for additional approval steps.
Can I pay with a credit card?
No, credit cards are not accepted. We only accept USDT on the TRON network (TRC-20) for top-ups. There are no cards, no PayPal, no subscriptions, and no invoices with VAT. You must use a compatible crypto wallet to send USDT to your account address.
Is there a monthly subscription fee?
No, there are no monthly plans or subscription fees. You pay only for the characters you generate. There are no seats, no minimum spend requirements, and no recurring charges. Your balance does not expire, and you retain funds until you choose to use them.
What happens if I run out of credits?
You must top up your balance to continue generating audio. Requests will fail if your balance is insufficient to cover the cost. Top-ups are processed automatically via USDT TRC-20 transfers. You can issue a new API key at any time if needed.
Do failed requests deduct from my balance?
No, failed requests are free. If synthesis fails, the charge is returned in full to your balance. You are only charged for successful audio generations. This ensures you do not pay for audio files that were not delivered due to server errors.
How do I check my current balance?
You can check your balance on the /get-api-key/ page after signing in. The dashboard displays your current balance, trial status, characters left, and usage history. This information is updated as you make requests or top up your account via USDT.
Is the trial credit valid forever?
No, the $0.50 trial credit is valid for 7 days. It is granted when you sign in with Google. Accounts created with email and password start without this trial credit. Unused trial credit expires after the 7-day period regardless of usage.
Can I mix currencies for top-ups?
No, we only accept USDT on the TRON network (TRC-20). There are no other currency options for top-ups. All balances are tracked in USD equivalent, but payments must be made in the specified cryptocurrency. No other tokens are supported.
Are there any hidden fees?
No, there are no hidden fees. The only cost is $0.01 per 1,000 characters. There are no monthly fees and no per-request fees. You pay exactly for the characters generated, and failed requests are free of charge.
How is the character count calculated?
Character count includes all text sent in the input field, including spaces, punctuation, and special characters. The limit is 4,096 characters per request. Each character is counted individually, so ensure your text fits within this limit to avoid request failures.
Do I need a new key for each project?
No, you can use the same API key for multiple projects. Each account has one active key, but you can issue a new key at any time, which invalidates the old one immediately. Use the same key across different applications for simplicity.
Cost calculator
Standard Character-Based Billing
The LinguaForge text-to-speech API utilizes a straightforward metered billing model based on character count. The standard rate is $0.01 per 1,000 characters, which equates to $10 per million characters processed. This structure ensures that costs scale linearly with your data generation volume without requiring complex tier calculations.
Rate per Thousand Characters
Each request is charged based on the total character count of the input text. The API charges $0.01 for every 1,000 characters used. This rate applies uniformly regardless of the voice selected or the audio format generated.
Space and Punctuation Inclusion
All characters within the input string are counted toward the total usage. This includes spaces, punctuation marks, and special symbols. Accurate character counting is essential for precise cost prediction in your ML training pipelines.
Predictable Cost Modeling
Since the rate is fixed, you can calculate exact project expenses before generating your audio corpus. This transparency allows for efficient budgeting when producing large-scale labeled datasets for machine learning applications.
Payment Infrastructure
Top-ups are processed exclusively via USDT on the TRON network (TRC-20). This infrastructure eliminates traditional payment gateways, reducing transaction friction and allowing for direct blockchain settlement. Users initiate transfers to the designated address, and the system monitors the network for confirmation.
Upon network confirmation, the corresponding USD value is automatically credited to your account balance. This automation ensures that funds are available for API requests within minutes of the transfer, maintaining continuous service availability without manual intervention or waiting for banking cycles.
USDT TRC-20 Transactions
Transactions utilize the TRC-20 standard on the TRON blockchain, ensuring low latency and minimal network fees compared to traditional fiat rails. This method supports high-throughput requirements typical of automated ML pipelines, allowing developers to integrate payment triggers directly into their data generation scripts without external payment processing delays.
- Supported currency: USDT (Tether) on TRON
- Standard: TRC-20
- Network: TRON mainnet
Automatic Balance Crediting
The system monitors the blockchain for incoming transactions and updates the account ledger automatically. Once the required number of confirmations is reached, the credit is applied immediately to your active key’s balance. This eliminates the need for support tickets to verify manual transfers, ensuring that your dataset generation jobs do not stall due to funding delays.
- Real-time network monitoring
- Instant balance update upon confirmation
- No manual verification required
Minimum Spend Requirement
Top-ups have a minimum deposit threshold of $5. This threshold is designed to cover network transaction costs and streamline the automated crediting process. There is no maximum limit, allowing for bulk purchases that align with large-scale corpus generation projects, ensuring consistent per-character pricing without volume-based tier changes.
- Minimum top-up: $5
- No maximum limit
- Transparent fee structure
Free Trial Allocation
New accounts authenticated via Google receive an immediate $0.50 trial credit. This allocation provides approximately 50,000 characters of synthesis capacity, allowing for immediate validation of the API's output quality without financial commitment.
The trial credit remains valid for a strict 7-day period from account creation. This timeframe is sufficient for developers to test voice presets and format options before deciding on a paid TRC-20 top-up.
Google Sign-In Bonus
Upon signing in with Google, the system automatically credits $0.50 to your balance. This instant grant enables immediate API usage for testing the tts-1 and tts-1-hd models.
7-Day Validity Period
The trial credit expires exactly 7 days after account creation, regardless of usage. Unused balance does not roll over. This ensures that test credits are used within a defined evaluation window.
Email Account Exclusion
Accounts created using only email and password do not receive the trial credit. To access the $0.50 bonus, you must authenticate via Google. Standard paid top-ups remain available for all account types.
- Google sign-in required for bonus
- No trial credit for email-only accounts
- Paid TRC-20 top-ups available to all
Cost Efficiency for Large Volumes
Generating labeled audio corpora at scale requires predictable unit economics. Our API charges a flat $0.01 per 1,000 characters regardless of volume. This consistent per-character rate ensures that your cost modeling remains accurate whether you are processing thousands or millions of files.
Consistent Per-Character Rate
Character counting includes spaces and punctuation, providing precise cost estimation for dataset creation. Since input text is limited to 4,096 characters per request, longer texts are split and joined automatically. This uniform pricing structure eliminates unexpected variances in your per-unit audio generation costs.
No Volume Tier Changes
The rate remains $0.01 per 1,000 characters across all usage levels. There are no volume tier changes or discounts applied as you scale your corpus. This linear cost model simplifies budgeting for machine learning pipelines without requiring complex negotiation or renegotiation of rates.
Zero Infrastructure Fees
There are no monthly platform fees or seat limits associated with high-volume usage. You pay only for the characters synthesized. This absence of fixed overhead costs makes the API particularly efficient for indie developers managing variable workloads without paying for idle capacity.

Refund and Failure Policy
The API returns a distinct status code when synthesis fails, ensuring that no characters are deducted from your balance. This applies to transient errors such as server timeouts or invalid voice parameters, guaranteeing that you pay only for successfully generated audio files.
Full Refund on Errors
When a request fails, the charge is returned in full immediately. The system logs the failure and adjusts your usage statistics accordingly, so your reported character count remains accurate.
No Charge for Failed Requests
You are not billed for any request that does not produce a valid audio file. This includes requests that exceed the 4096-character limit, ensuring predictable cost modeling for your dataset generation pipelines.
Billing Transparency
Your dashboard displays the exact balance, characters consumed, and any failed requests that were refunded. This clarity allows you to audit your spending accurately without hidden fees or prorated charges for partial generations.
Comparison with Traditional Licensing
LinguaForge replaces traditional TTS licensing models with a pure per-character billing structure, eliminating the complexity of seat management and recurring subscription fees. This approach aligns costs directly with usage volume, ensuring that enterprises pay only for the audio generated during active training or inference pipelines.
Our API supports concurrent requests from any number of developers or servers. This flexibility is essential for distributed machine learning workflows where multiple workers generate labeled audio datasets simultaneously without additional overhead.
Per-Character vs. Subscription
Traditional licensing often requires monthly subscriptions that may not match actual usage patterns, leading to overpayment during low-activity periods or capacity limits during spikes. Our model charges $0.01 per 1,000 characters, providing predictable cost modeling for budgeting large-scale corpus creation without unexpected tier upgrades.
- Pay only for actual text length, including spaces and punctuation
- No fixed monthly fee regardless of API call volume
- Ideal for variable workloads typical in ML dataset generation
No Seat Limits
Legacy enterprise plans often restrict the number of active user accounts or API keys, requiring additional fees for each new developer or service instance. LinguaForge imposes no seat limits, allowing teams to issue multiple keys for different microservices or testing environments without additional per-seat costs.
- Unlimited API keys per account
- Support for distributed worker nodes
- No penalty for team expansion
No Monthly Commitments
Many competitors enforce annual contracts or minimum monthly spend requirements, creating financial rigidity for agile development teams. Our top-up model allows immediate balance adjustments via USDT TRC-20, with no minimum spend and no expiring credits, ensuring funds remain available for long-term projects without renewal friction.
- No annual contracts or auto-renewal
- Balance does not expire
- Immediate credit via automated TRC-20 transfers
Example Cost Scenarios
Pricing remains constant regardless of volume, allowing precise budget forecasting for ML pipelines. Each character, including spaces and punctuation, incurs a uniform charge of $0.01 per 1,000 characters.
This linear model eliminates hidden fees, making cost estimation straightforward for both small experiments and large-scale data generation projects.
Small Dataset Generation
Generating a 50,000-character training set costs $0.50, equivalent to the free trial credit available via Google sign-in. This allows immediate validation of audio quality without upfront payment.
- Input: 50,000 characters
- Cost: $0.50
- Output: MP3/WAV audio files
Large-Scale Corpus Creation
For a 1 million-character corpus, the base cost is $10.00. Users can top up via USDT (TRC-20) with a minimum of $5, ensuring continuous generation without interruption.
With 60 requests per minute, generating large datasets requires splitting text into 4,096-character chunks. Processing time approximates real-time, taking roughly 16.5 hours for 1 million characters under steady load, though requests may queue under load.
- Input: 1,000,000 characters
- Base Cost: $10.00
- Min Top-up: $5.00 (USDT TRC-20)
Stress Testing Budgeting
Rate limits allow 60 requests per minute per key. Under heavy load, requests queue sequentially, potentially extending total generation time beyond real-time estimates.
Budgeting should account for this queueing effect. Failed requests incur no charge, ensuring that infrastructure retries do not inflate costs unexpectedly.
- Limit: 60 requests/minute
- Retry Cost: $0.00 (if failed)
- Queue Impact: Variable latency
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.
Sources and further reading
- Wikipedia
- Python documentation
- W3C WAI: making audio and video accessible
- LibriVox public-domain audiobooks
- OpenAI API reference: create speech
- openai-python on GitHub
- IETF RFC 7845: Ogg Opus
- ISO 639 language codes
- Creative Commons Attribution 4.0
- LibriTTS-R corpus (OpenSLR 141)
- TRON developer hub: TRC-20
- LibriTTS-R paper (arXiv 2305.18802)
