Text to Speech API Pricing and Character Costs

By SynthEval editorial team · Updated 2026-11-18

This page details the character-based pricing for SynthEval’s text-to-speech API, designed for startups generating labeled audio datasets. The rate is $10 per 1 million characters, with spaces and punctuation counted equally. There is no minimum spend or monthly commitment, making it easy to scale usage up or down based on your ML pipeline needs.

You can start with a $0.50 free trial valid for 7 days if you sign in with Google. Payments are processed via USDT on the TRON network (TRC-20), with a $5 minimum top-up. Failed requests are free, and your balance never expires. Track usage via the dashboard and manage keys instantly.

Text to Speech API Pricing and Character Costs

Frequently asked questions

How much does the text to speech API cost?

The API costs $10 per 1 million characters, which equals $0.01 per 1,000 characters. This flat rate applies to all models and voices without tiered pricing structures.

Do spaces count as characters in billing?

Yes, spaces and punctuation count as characters in billing. The character count includes every character in your input text, so ensure your input length calculations account for whitespace.

Is there a monthly subscription fee?

No, there are no monthly plans or subscriptions. You pay only for the characters you generate, with no recurring fees or minimum monthly spend requirements for your account.

What payment methods are accepted?

We accept USDT on the TRON network (TRC-20) only. We do not accept credit cards, PayPal, or invoices. There are no subscriptions or seats, only pay-as-you-go top-ups.

How long does USDT top-up take?

Top-ups are credited automatically within minutes after the transfer is confirmed on the blockchain. The minimum top-up amount is $5, and the process is fully automated without manual review.

Do I pay for failed requests?

No, failed requests are free. If synthesis fails, the charge is returned in full to your balance. You only pay for successfully generated audio files delivered to your client.

Is there a minimum spend requirement?

No, there is no minimum spend requirement for your account. You can top up any amount above the $5 minimum for USDT transactions and use the balance as needed.

Does my balance expire?

No, your balance does not expire. Unlike some services that impose time limits on credits, your remaining credit remains available indefinitely until you choose to generate more audio or close the account.

How do I check my remaining characters?

You can check your remaining characters and balance on the /get-api-key/ dashboard page. The interface displays your current balance, trial status, and usage statistics in real time for your account.

Can I get a refund for unused credit?

The product facts do not specify a refund policy for unused credit. Refunds are only explicitly mentioned for failed requests, where the charge is returned in full.

Are there volume discounts for large datasets?

No, there are no volume discounts. The rate remains $10 per 1 million characters regardless of volume. Pricing is flat and consistent for all customers, from small tests to large datasets.

Why is USDT the only payment method?

We use USDT on TRON to avoid traditional payment processing fees and enable instant, borderless settlements. This allows us to offer no monthly plans, no subscriptions, and no VAT invoices.

How is the $0.50 trial credit applied?

The $0.50 trial credit is applied automatically when you sign in with Google. It provides about 50,000 characters valid for 7 days. Email/password accounts do not receive this trial credit.

What happens if I run out of credits mid-request?

Requests are processed one after another. If you run out of credits, subsequent requests may wait in a queue or fail. Failed requests are free, but successful ones require sufficient balance.

Cost calculator

Estimated cost: $2.50

Spaces and punctuation count as characters. $10 per 1 million characters. Failed requests are free.

Standard Pay-As-You-Go Model

Sound wave illustration for: Standard Pay-As-You-Go Model

SynthEval charges a flat rate of $10 for every 1 million characters processed. This straightforward model eliminates complex tiering, making cost forecasting simple for teams generating labeled audio datasets at scale. You only pay for the text that enters the engine, with a clear line between input and output costs.

Rate of $10 per 1 million characters

Rate of $10 per 1 million characters

The base cost is fixed at $10 per million characters. This applies to both tts-1 and tts-1-hd models, which share the same engine and pricing structure. For budgeting purposes, this equates to $0.01 per 1,000 characters, providing predictable unit economics for ML training pipelines.

  • Flat rate: $10 / 1M characters
  • Applies to tts-1 and tts-1-hd
  • $0.01 per 1,000 characters

Spaces and punctuation count as characters

Character counting includes spaces and punctuation marks, ensuring accurate billing for every element of your input text. Long texts are split into parts of up to 4096 characters and joined as audio files. This means precise control over volume costs, as every comma and newline contributes to the total character count.

No minimum spend or monthly commitment

There are no monthly fees or required minimum spends. You can pause or resume generation at any time without penalty. This flexibility allows startups to scale usage up or down based on project needs, paying only for the synthetic speech data actually generated during the billing period.

Free Trial and Onboarding

Sound wave illustration for: Free Trial and Onboarding

New users can start generating synthetic speech immediately without entering payment details. By signing in with Google, the account is automatically credited with $0.50, providing approximately 50,000 characters for testing. This allows teams to verify voice quality and API integration.

$0.50 credit for Google sign-ups

The trial credit is applied instantly upon authentication. This amount covers roughly 50,000 characters, allowing you to test multiple voices and formats to ensure the synthetic audio meets your dataset requirements.

Trial valid for 7 days

The trial credit remains active for 7 days from account creation. If not used within this window, the credit expires, but no charges are incurred. This window is sufficient for initial integration testing and format validation.

Email/password accounts start with zero credit

Accounts created using email and password do not receive the trial bonus. These users begin with a zero balance and must top up via USDT on the TRON network to start generating audio.

Free Trial and Onboarding

Payment Methods and Currency

Sound wave illustration for: Payment Methods and Currency

To fund your SynthEval account, we accept a single currency method: USDT on the TRON network (TRC-20). This approach eliminates traditional banking delays and credit card processing fees, ensuring your credits are available for immediate dataset generation.

USDT on TRON network (TRC-20) only

We do not support PayPal, credit cards, or traditional wire transfers. All top-ups must be sent as TRC-20 tokens to the address provided in your dashboard. This single-currency model simplifies accounting for international ML startups.

  • No fiat currency conversion fees
  • Single supported token standard
  • Direct blockchain settlement

Minimum top-up amount of $5

The minimum deposit is $5.00. There is no monthly commitment or minimum spend requirement. This low threshold allows small teams to test the API with a minimal financial outlay before scaling up for large-scale audio dataset production.

Automatic credit within minutes of transfer

Once the blockchain confirms your transfer, credits are added to your account automatically within minutes. You can immediately begin generating synthetic speech without waiting for manual finance team approval or invoice processing.

Cost Efficiency for Large Datasets

Sound wave illustration for: Cost Efficiency for Large Datasets

Generating labeled audio datasets for machine learning requires consistent, low-cost synthesis. Our flat rate of $10 per 1 million characters allows startups to predict expenses accurately, whether producing short phrases or large-scale corpora. This structure eliminates surprise charges, ensuring your budget remains stable during intensive data collection phases.

Competitive rate for bulk audio generation

The per-character model scales linearly, making it ideal for high-volume ML pipelines. You pay only for the text processed, with no minimum spend or monthly commitment. This approach supports rapid iteration without the overhead of tiered plans.

No extra fees for format selection

Choose from mp3, wav, opus, flac, or pcm without incurring additional costs. Format selection is purely a technical decision for your dataset requirements, not a pricing factor. This consistency simplifies infrastructure planning for diverse model training needs.

No charges for failed synthesis requests

If synthesis fails, the charge is returned in full. You are not billed for wasted compute or errors, reducing the risk of generating unusable data. This ensures that your character count reflects only successful audio production.

Cost Efficiency for Large Datasets

Balance Management

Sound wave illustration for: Balance Management

Manage your synthetic speech credits with full transparency. Your balance never expires, allowing you to pause large-scale dataset generation projects without losing funds. This structure supports flexible scaling for ML pipelines where audio generation needs may fluctuate between intensive training phases and quieter periods.

Balance does not expire

Credits remain available indefinitely until used. This eliminates the pressure of monthly usage quotas, letting you align audio synthesis with your specific data collection cycles without worrying about credit forfeiture.

Real-time usage tracking in dashboard

The /get-api-key/ dashboard displays your current balance, characters remaining, and usage history. You can monitor exactly how many characters have been processed, ensuring precise budget control for your ML projects.

Instant key revocation and reissuance

You maintain one active API key per account. If needed, you can issue a new key immediately, causing the old key to stop working at once. This allows quick rotation without downtime or complex permission management.

Volume Discounts and Scaling

Sound wave illustration for: Volume Discounts and Scaling

For synthetic speech dataset generation, volume does not alter the unit rate. Whether you are processing 50,000 characters for a small pilot or 10 million characters for large-scale model training, the price remains strictly $10 per 1 million characters. This flat structure eliminates the complexity of negotiating volume tiers, allowing startups to scale their audio data production linearly with their budget.

Flat rate regardless of volume

The API charges a uniform rate for every character processed, irrespective of total volume. This consistency ensures predictable costs, regardless of whether you are generating short phrases or lengthy narratives. There are no hidden penalties for high-throughput bursts, provided you stay within the 60 requests-per-minute rate limit per key.

No tiered pricing structures

Unlike traditional cloud providers that require you to jump between pricing tiers based on usage thresholds, our model applies a single rate across all volumes. You do not need to upgrade to a 'Pro' or 'Enterprise' plan to access better rates; the standard pay-as-you-go structure applies equally to small experiments and industrial-scale dataset creation.

Predictable costs for project budgeting

Since the cost is directly tied to character count, you can precisely calculate the expense of any audio dataset before generation. For example, a 1-million-character dataset will always cost exactly $10, plus any applicable network fees. This predictability simplifies financial planning for ML projects, where audio data volume is often estimated weeks in advance.

Volume Discounts and Scaling

What Is Not Included

Sound wave illustration for: What Is Not Included

We operate on a pure API model, so you will not receive traditional VAT invoices or paper receipts. Billing records are available directly in your dashboard, but formal tax documentation is not generated automatically.

Our payment infrastructure is streamlined for speed. We do not support PayPal or credit card processing, eliminating the friction of recurring billing cycles and subscription management for a one-time utility service.

No VAT invoices or receipts

Invoices are not provided. All transaction history is visible in your account dashboard for your accounting needs.

No PayPal or credit card options

We do not accept PayPal or credit cards. All payments are made via USDT on the TRON network (TRC-20). This ensures low transaction fees and instant settlement without banking delays.

The minimum top-up amount is $5. Credits are added automatically within minutes of confirming the transfer.

No enterprise contracts

We do not offer enterprise contracts, SLAs, or dedicated support lines. Our pricing is flat and transparent: $10 per million characters regardless of volume. There are no tiered pricing structures or volume discounts to negotiate.

This model allows startups to scale usage predictably without negotiating custom terms. Support is handled via email only.

  • No SLAs or uptime guarantees
  • No dedicated account managers
  • No custom volume discounts

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. Our synthetic speech datasets help startups train smarter AI models without expensive voice talent.

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. LibriVox public-domain audiobooks
  3. openai-python on GitHub
  4. FFmpeg documentation
  5. LibriTTS-R corpus (OpenSLR 141)
  6. Tether (USDT)
  7. OpenAI API reference: create speech
  8. IETF RFC 7845: Ogg Opus
  9. Creative Commons Attribution 4.0
  10. TRON developer hub: TRC-20
  11. curl documentation
  12. W3C WAI: making audio and video accessible

More on this site