API for News Publishers: Fast Audio Generation

By PressSynth editorial team · Updated 2026-11-19

This page provides a practical guide to PressSynth, a text-to-speech API designed for news publishers who need reliable audio generation without enterprise contracts or SLAs. You will find clear instructions on integrating our Python-friendly API into your digital news workflows, covering everything from voice selection to rate limits.

We detail the technical specifications, transparent character-based pricing, and data privacy practices that support scalable audio production for newsletters and scripts. Whether you are generating neutral news delivery or documentary-style features, this guide helps you implement the solution efficiently.

API for News Publishers: Fast Audio Generation

Frequently asked questions

Is this API compatible with the official OpenAI Python SDK?

Yes. The endpoint accepts the same JSON body shape as the OpenAI speech endpoint. You can use the official OpenAI SDKs by changing the base_url to https://api.98wi.info/v1 and providing your PressSynth API key. This allows for easy integration into existing Python workflows without requiring a custom client library.

How many characters can I send in a single request?

You can send up to 4096 characters per request. For longer texts, split the input into parts of this size and join the resulting audio files. This limit applies to each individual API call, ensuring consistent processing times and predictable billing based on the character count.

What happens if a request fails during generation?

Failed requests are free. If synthesis fails, the charge is returned in full. You are only billed for successful audio generation. This ensures you do not pay for incomplete or erroneous outputs, maintaining cost efficiency for your news audio production pipeline.

Do you offer voice cloning for specific news anchors?

No. Custom voice cloning is not offered. You cannot upload your own voice or design custom voices. The API provides eight preset voices based on licensed recordings. If you need a specific anchor's voice, you must select the closest matching preset from the available options.

Are there any hidden fees for using the API?

No. Pricing is transparent at $0.01 per 1,000 characters. Failed requests are free. There are no monthly plans, seats, or minimum spends. You pay only for the characters generated. Top-ups via USDT on the TRON network are the only payment method, with no hidden subscription costs.

Which voices are best for reading news articles?

Atlas is designed as a neutral, news-oriented voice. Milo offers a friendly, conversational tone suitable for interviews or lighter news. For serious reporting, Atlas provides a standard broadcast quality. Other voices like Sage (calm, instructional) may also work well depending on the specific content style.

Can I use this API for commercial news broadcasts?

Yes. The API is built for publishers and commercial use. Generated audio files are returned directly in the response and are not stored by the service. You retain the right to use the audio for digital news, newsletters, and broadcasts without additional licensing fees per file.

How is the language detected if I don't specify it?

The language is guessed from the script. Latin script text defaults to English. For non-English Latin-script text, you should explicitly set the language parameter. If not set, the system assumes English, which may affect pronunciation accuracy for other languages using Latin characters.

What audio formats are available for download?

Available formats are mp3, wav, opus, flac, and pcm. You specify the desired format in the request body using the response_format field. These formats cover most broadcast and digital distribution needs, from compressed web audio to high-quality uncompressed files for professional editing.

Do I need to pay for each voice or is it included?

All eight voices are included in the standard price. There is no extra cost for using different voices. The price is based on character count ($0.01 per 1,000 characters), regardless of which voice you select from the available presets like Aria, Nova, or Atlas.

Is there a limit on how many requests I can make per minute?

Yes, the limit is 60 requests per minute per key. Before the first top-up, the limit is 10 requests per minute. This rate limit applies to all API calls. If you exceed the limit, requests may be queued or rejected until the window resets.

How long is my API key valid?

Your API key does not expire. It remains valid as long as your account is active. You can issue a new key at any time, which immediately invalidates the old one. This allows you to rotate keys for security without losing access to your balance.

Can I upgrade my key if I need higher throughput?

You can issue a new API key at any time. The rate limit is 60 requests per minute per key. Issuing a new key provides a separate limit. To increase throughput, you may need multiple keys.

Do you provide SSL certificates for secure connections?

The API endpoint uses standard HTTPS (https://api.98wi.info/v1). SSL/TLS is handled by the infrastructure provider. You do not need to manage certificates yourself. Connections are secured using standard web protocols, ensuring your API keys and data are transmitted securely.

First request

curl

curl https://api.98wi.info/v1/audio/speech \
  -H "Authorization: Bearer $TTS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"tts-1","voice":"aria","input":"Hello from the API.","response_format":"mp3"}' \
  --output hello.mp3

Why publishers choose this API

Sound wave illustration for: Why publishers choose this API

Publishers seeking reliable audio generation without enterprise overhead find this API aligns with operational needs. The service provides a straightforward Python-friendly interface for converting articles and scripts into broadcast-quality audio, eliminating the need for complex negotiations or long-term commitments.

No enterprise contracts or SLAs

No enterprise contracts or SLAs

PressSynth operates without enterprise contracts or SLAs, offering a flexible solution for newsrooms that prefer simplicity over guaranteed uptime metrics. This approach allows publishers to integrate audio generation directly into their workflows without administrative burden or minimum spend requirements.

Simple Python integration

The endpoint follows the standard OpenAI shape, allowing teams to use the official OpenAI SDK by simply changing the base URL. This compatibility ensures that developers familiar with common TTS libraries can deploy audio generation for newsletters and digital stories with minimal code changes.

  • Standard JSON request body
  • Compatible with OpenAI SDKs
  • Easy base URL configuration

Transparent character pricing

Pricing is calculated at $0.01 per 1,000 characters, with spaces and punctuation counting toward the total. Failed requests are refunded in full, ensuring you only pay for successful synthesis. This model scales naturally with your content volume, from daily briefs to long-form documentaries.

Supported voices for news

Sound wave illustration for: Supported voices for news

PressSynth offers eight voices suitable for news publishing workflows. Atlas delivers a neutral delivery suitable for breaking news updates. Orion provides a deep, documentary-style tone suitable for long-form features. Sage offers a calm, instructional voice perfect for explainers.

These presets are built on licensed recordings and include inaudible synthetic watermarks. All voices support the full range of 23 languages, allowing publishers to localize content without hiring new talent for each region.

Atlas for neutral news delivery

Atlas provides a clear, neutral voice suitable for broadcast news delivery. It is designed for clarity and consistency, making it the default choice for daily news briefs and headline summaries.

The voice maintains a steady pace and neutral tone, ensuring that factual content is delivered without distracting emotional inflection. This makes it ideal for high-volume news feeds where consistency matters more than character.

  • Ideal for breaking news and headlines
  • Neutral tone ensures factual clarity
  • Consistent performance across 23 languages

Orion for documentary-style features

Orion offers a deep, resonant voice suited for documentary-style features and long-form journalism. Its tone adds gravity and depth to investigative reports or historical context segments.

Use Orion when you want to elevate standard text into a more immersive audio experience. The voice works well for serialized content or special reports that require a distinct narrative presence.

  • Deep, resonant tone for features
  • Ideal for investigative journalism
  • Enhances long-form narrative content

Sage for instructional content

Sage delivers a calm, instructional voice perfect for educational news segments or how-to guides. It is clear and approachable, making complex topics easier to digest for listeners.

Sage is best used for explanatory content, weather updates, or traffic reports where clarity and a gentle tone help listeners retain information without feeling rushed or overwhelmed.

  • Calm and approachable tone
  • Ideal for explanations and guides
  • Clear articulation for complex topics

Language support for global reach

Sound wave illustration for: Language support for global reach

PressSynth supports 23 languages, including English, Spanish, German, French, Arabic, Chinese, Japanese, and Hindi. Every voice can speak any of these languages, allowing publishers to localize content without managing multiple voice assets. The API covers major global markets while maintaining consistent audio quality across all supported tongues.

23 languages supported

The API handles text in Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. If a language is not listed, it is not supported.

  • Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish

Automatic language detection

If the language field is omitted, the system guesses the language from the text script. Latin-script text defaults to English. This works well for single-language articles but may misidentify mixed-language content.

Explicit language setting for accuracy

For non-English Latin-script text or mixed scripts, always set the language parameter to the ISO code. This ensures the correct pronunciation and intonation patterns are applied by the chosen voice preset.

Example: "language": "de" for German text. This prevents the engine from defaulting to English pronunciation for words like "news" or "data".

Technical specifications

Sound wave illustration for: Technical specifications

PressSynth uses a standard HTTP endpoint compatible with the OpenAI SDK, allowing publishers to integrate audio generation into their existing Python workflows with minimal code changes. By setting the base URL to https://api.98wi.info/v1, you can use standard libraries like openai-python to send text and receive audio files directly.

OpenAI-compatible endpoint

Send a POST request to /v1/audio/speech with a JSON body containing the text, voice selection, and desired format. The API accepts the same request shape as the OpenAI speech endpoint, ensuring that official SDKs work seamlessly by simply adjusting the base URL.

Standard audio formats

Generate audio in mp3, wav, opus, flac, or pcm formats to suit different delivery channels. MP3, WAV, and FLAC are available formats. PCM is available for raw audio processing pipelines.

Rate limits and queueing

Keys are limited to 60 requests per minute, with a lower limit of 10 requests per minute for new accounts before the first top-up. Requests are processed sequentially, so under heavy load, synthesis may wait in a queue. Typical generation speed is roughly real-time, producing about 25 characters of text per second of audio.

Pricing model

Sound wave illustration for: Pricing model

PressSynth charges $0.01 per 1,000 characters, including spaces and punctuation. Failed synthesis requests are refunded in full, ensuring you only pay for delivered audio. The balance never expires, and top-ups via USDT (TRC-20) credit automatically within minutes.

There are no monthly minimums, subscriptions, or enterprise contracts. You pay only for what you generate, making it ideal for publishers with variable output volumes. Accounts start with $0.50 trial credit for 7 days when signing in with Google.

$0.01 per 1,000 characters

The standard rate is $10 per 1 million characters. Each character, including spaces and punctuation, counts toward this total. If synthesis fails, the charge is returned immediately, so you are not billed for errors.

  • $0.01 per 1,000 characters
  • Failed requests are free
  • Balance does not expire

No monthly minimums

There are no monthly plans, seats, or minimum spend requirements. You can top up with as little as $5 via USDT on the TRON network. Credits are added automatically within minutes of transfer confirmation.

Free trial credit available

Signing in with Google provides $0.50 of credit, valid for 7 days. This covers approximately 50,000 characters. Accounts created with email and password do not receive this trial credit.

This allows you to test the API integration before committing to larger volumes of audio generation for your newsletters or news feeds.

  • $0.50 credit for Google sign-ins
  • Valid for 7 days
  • Approx. 50,000 characters
Pricing model

Data and privacy

Sound wave illustration for: Data and privacy

For publishers processing article drafts, input text is used strictly to produce the audio file and is not used for training the model. This ensures your editorial content remains confidential during synthesis.

Input text not used for training

Your article text is processed solely to generate the audio response. It is not ingested into the training dataset, preserving the exclusivity of your news content.

Audio not stored after generation

Generated audio files are returned in the API response and are not permanently stored on our servers. Once you receive the file, the temporary asset is cleared.

Billing metadata retention

We retain request metadata, including timestamp, character count, and status, solely for accurate billing and usage verification. This data does not include the content of your articles.

Integration workflow

Sound wave illustration for: Integration workflow

PressSynth integrates with existing publisher workflows using standard tools. Begin by creating an account at /get-api-key/ to retrieve your unique API key. This key authenticates your requests and tracks usage against your balance.

Sign up and get API key

Navigate to the registration page to create your account. You can sign in using Google or an email/password combination. Once registered, your dashboard displays your API key, current balance, and usage statistics.

  • Google or email/password authentication
  • Immediate access to API key
  • Real-time balance tracking

Install OpenAI SDK

The API is compatible with the official OpenAI Python SDK. Install the library using your standard package manager. This allows you to use familiar methods for sending requests and handling responses.

Configure base URL

Update your client configuration to point to our endpoint. Set the base URL to https://api.98wi.info/v1. This ensures all requests are routed correctly to the PressSynth engine while maintaining standard request shapes.

Performance expectations

Sound wave illustration for: Performance expectations

Generation speed is roughly real-time, processing approximately 25 characters of text per second of audio output. While long-form articles generate at this steady pace, short phrases may take a few seconds to complete due to processing overhead.

Under heavy load, requests are processed sequentially, meaning a request may wait in a queue. This ensures fair resource distribution but can increase latency during peak usage periods.

Real-time generation speed

Generation speed is roughly real-time, processing approximately 25 characters of text per second of audio output. While long-form articles generate at this steady pace, short phrases may take a few seconds to complete due to processing overhead.

Queue handling under load

Under heavy load, requests are processed sequentially, meaning a request may wait in a queue. This ensures fair resource distribution but can increase latency during peak usage periods.

Character limit per request

Each request supports up to 4096 characters. For articles exceeding this limit, split the text into parts and join the resulting audio files.

Failed requests do not consume credit, ensuring you only pay for successfully generated audio.

Support and contact

Sound wave illustration for: Support and contact

PressSynth provides support via email for API integration or billing questions. Our team reviews inquiries during business hours and responds with specific, actionable guidance tailored to your newsroom’s technical stack.

We do not offer live chat, phone lines, or dedicated account managers, keeping our structure lean and cost-effective for independent and mid-sized publishers who prefer asynchronous communication.

For immediate technical reference, our documentation covers endpoint specifications, SDK examples, and troubleshooting steps for common integration scenarios.

Email support only

Send your questions to [email protected]. Include your API key prefix and a description of the issue, such as rate limit errors or voice selection questions. We prioritize clear, written responses to ensure reproducibility.

No live chat or phone

There is no real-time chat widget or toll-free number. This approach reduces overhead, allowing us to maintain transparent character-based pricing without enterprise support fees. All support interactions are logged for reference.

Documentation resources

Our docs detail the OpenAI-compatible endpoint, voice presets, and language codes. Use the /docs/ page for code snippets, curl examples, and rate limit details to resolve most integration questions independently.

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. The service provides eight preset voices based on licensed recordings from LibriTTS-R and LibriVox.

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. LibriTTS-R corpus (OpenSLR 141)
  3. Python documentation
  4. Creative Commons Attribution 4.0
  5. OpenAI API reference: create speech
  6. MDN: Web Speech API
  7. TRON developer hub: TRC-20
  8. Tether (USDT)
  9. openai-python on GitHub
  10. FFmpeg documentation
  11. openai-node on GitHub
  12. IETF RFC 7845: Ogg Opus

More on this site