Frequently asked questions
Which TTS API has the most languages?
Q: Which TTS API has the most languages? A: PressSynth supports 23 languages, including Arabic, Chinese, English, French, German, Hindi, Japanese, Korean, and Spanish. Any voice can speak any of these languages. We do not support Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, Bengali, or other languages not in our list.
Is PressSynth better than OpenAI for news?
PressSynth uses the same request shape as OpenAI’s speech endpoint, allowing easy integration. Our pricing is $0.01 per 1,000 characters, with no monthly plans or enterprise contracts. We provide 8 distinct voices suitable for digital news and newsletters without hidden fees. Unlike OpenAI, we are not affiliated with them, offering a direct, publisher-focused API for scalable audio generation.
Do these APIs support voice cloning?
PressSynth does not support custom voice cloning, voice design, or uploading your own voice. Voices are based on licensed recordings from LibriTTS-R and LibriVox. We offer 8 preset voices, such as atlas for neutral news reading, but do not provide emotion tags or phoneme dictionaries.
What is the cheapest TTS API for publishers?
PressSynth charges $10 per 1 million characters ($0.01 per 1,000 characters). Failed requests are free. We have no monthly plans, no seats, and no minimum spend. Top-ups via USDT (TRC-20) are credited automatically. This flat rate applies to all text lengths up to 4,096 characters.
Which API has the best Python documentation?
PressSynth uses the same request shape as OpenAI, so the official OpenAI SDKs work by changing the base_url to https://api.98wi.info/v1. You can also use standard Python libraries to send POST requests. We provide a ready curl example and clear documentation on our /docs/ page.
Can I use these APIs for commercial news?
Yes. PressSynth voices are based on licensed recordings (LibriTTS-R CC BY 4.0) and public-domain audiobook readings. Generated audio files are returned in the response and not stored. Input text is used only for synthesis and not for training. You can use the output for commercial news broadcasts.
Do any APIs offer a free trial?
PressSynth offers a $0.50 credit (about 50,000 characters) valid for 7 days when you sign in with Google. Accounts created with email and password start without trial credit. Failed requests are also free, allowing you to test the API without charge if synthesis fails.
How does pricing compare between providers?
PressSynth charges $0.01 per 1,000 characters. We have no monthly plans, no subscriptions, and no invoices with VAT. Top-ups are done via USDT on the TRON network. Unlike many providers, we do not charge for failed requests, and our balance does not expire.
Is OpenAI's TTS API compatible with PressSynth?
Yes. PressSynth uses the same request shape as the OpenAI speech endpoint. You can use the official OpenAI SDKs by changing the base_url to https://api.98wi.info/v1. The endpoint is POST https://api.98wi.info/v1/audio/speech, accepting the same JSON body structure with model, input, voice, and speed fields.
Which voice is best for news reading?
Atlas is our neutral voice, optimized for news reading. Sage provides a calm, instructional tone, while Felix suits energetic promos. All eight voices support 23 languages and can be selected via the /v1/voices endpoint. Since any voice can speak any supported language, choose the preset that best matches your publication’s specific style.
Do I need an enterprise contract for high volume?
No. PressSynth has no enterprise contracts, no seats, and no minimum spend. You can process unlimited requests as long as you have sufficient balance. Rate limits are 60 requests per minute per key. You can issue a new key if needed, stopping the old one immediately.
Are there any hidden costs in these APIs?
PressSynth has no hidden costs. Pricing is strictly $0.01 per 1,000 characters. Failed requests are free. There are no monthly plans, no subscription fees, and no VAT invoices. Top-ups via USDT are credited automatically. Balance does not expire, and you pay only for what you use.
Can I switch APIs without changing my code?
If you use the OpenAI SDK, you can switch to PressSynth by changing the base_url to https://api.98wi.info/v1. The request shape is identical. Ensure you update your Authorization header with your PressSynth key. This allows easy migration without rewriting your Python integration logic.
Which API has the lowest latency?
PressSynth processes requests sequentially. Typical speed is roughly real time (25 characters per second); short phrases take a few seconds. Under load, requests may wait in a queue. We do not offer streaming or WebSocket support, so latency depends on queue depth and text length. With 60 requests per minute, throughput is limited by this single-file processing model.
Evaluation criteria
To evaluate PressSynth for newsroom workflows, we prioritize three core metrics: the breadth of supported languages, the natural quality of the eight voice presets, and seamless compatibility with existing Python infrastructure.
This approach ensures the API meets the specific demands of digital publishers who need reliable, scalable audio generation for newsletters and broadcasts without enterprise overhead.
Language support breadth
The API supports 23 languages, covering major global markets including English, Spanish, German, French, Japanese, and Arabic. Any voice can speak any of these languages, though setting the language parameter is recommended for non-English Latin-script text to avoid default English guessing.
Notably, Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, and Bengali are not supported, which may limit usage for specific regional publications.
- 23 supported languages
- Any voice can speak any supported language
- Automatic language guessing for Latin script defaults to English
Voice variety and quality
PressSynth offers eight preset voices: four female (Aria, Nova, Luna, Sage) and four male (Felix, Milo, Atlas, Orion). Each voice is tuned for specific tones, from warm narration to neutral news delivery, ensuring consistency across long articles and short alerts.
Every generated audio file includes an inaudible watermark marking it as synthetic speech, maintaining transparency for broadcast standards.
Python SDK compatibility
The API follows the OpenAI speech endpoint structure, allowing direct use of official OpenAI Python SDKs by simply changing the base_url to https://api.98wi.info/v1. This eliminates the need for custom wrappers or learning new libraries.
Authentication is handled via a standard Bearer token, making integration straightforward for developers familiar with HTTP requests or existing OpenAI integrations.
PressSynth API overview
PressSynth is a cloud-based text-to-speech API designed for publishers who need reliable audio generation for digital news and newsletters. By sending text to a simple endpoint, users receive an audio file ready for broadcast or newsletter distribution. The system processes requests sequentially, handling longer texts by splitting them into 4,096-character segments and joining the resulting audio files.
8 preset voices
The service provides eight preset voices based on licensed recordings from LibriTTS-R and LibriVox. Female options include Aria (warm narrator), Nova (upbeat), Luna (storytelling), and Sage (instructional). Male options feature Felix (energetic), Milo (conversational), Atlas (neutral news), and Orion (deep documentary). Each voice is tuned for specific tones, from warm narration to neutral news delivery, ensuring consistency across long articles and short alerts.
- Aria: warm, clear narrator
- Nova: bright, upbeat
- Luna: soft, storytelling
- Sage: calm, instructional
- Felix: energetic, promo
- Milo: friendly, conversational
- Atlas: neutral, news
- Orion: deep, documentary
23 languages supported
The API supports 23 languages, including Arabic, Chinese, English, French, German, Japanese, Spanish, and many others. Any voice can speak any of these languages. If the language parameter is omitted, the system guesses the language from the script, defaulting to English for Latin text. For non-English Latin-script text, users should explicitly set the language code.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
$0.01 per 1,000 characters
Pricing is straightforward at $10 per million characters, equating to $0.01 per 1,000 characters. Spaces and punctuation count toward the total. Failed requests are free, with charges returned in full if synthesis fails. The API follows the OpenAI speech endpoint structure, allowing direct use of official OpenAI Python SDKs by simply changing the base_url to https://api.98wi.info/v1.
Comparison with major brands
PressSynth competes directly with established providers by offering identical OpenAI-compatible endpoints, allowing newsrooms to switch providers without code changes. Unlike ElevenLabs, which specializes in deep voice cloning, PressSynth focuses on stable, licensed presets suitable for consistent broadcast news. Google Cloud TTS offers extensive neural models but requires complex integration; PressSynth simplifies this with a single API key and straightforward Python SDK support.
This comparison highlights PressSynth’s practical advantages for publishers seeking reliable, scalable audio generation without enterprise-level overhead or proprietary lock-in. The platform’s fixed pricing and multi-language support make it a viable alternative for newsletters and digital articles.
OpenAI TTS feature set
PressSynth uses the exact same request shape as OpenAI’s speech endpoint. You can swap the base URL in the official OpenAI Python SDK to use PressSynth voices immediately. This ensures compatibility with existing codebases used by many newsrooms.
The model supports tts-1 and tts-1-hd, providing high-quality audio at a fraction of the cost per character. This parity allows publishers to test quality without rewriting their audio generation pipelines.
- Identical JSON request structure
- Works with official OpenAI SDK
- No code refactoring required
ElevenLabs voice cloning
While ElevenLabs excels in custom voice cloning, PressSynth offers eight preset voices based on licensed LibriTTS-R recordings. This approach ensures consistent, broadcast-ready audio without the variability of cloned voices.
For news publishers who need reliable, standardized narration for articles and scripts, preset voices provide uniformity across different pieces of content. Cloning is not offered, keeping the service simple and predictable.
The lack of cloning does not hinder news applications where voice consistency across multiple articles is preferred over unique character voices.
- Eight licensed presets
- No custom cloning
- Consistent broadcast quality
Google Cloud TTS integration
Google Cloud TTS offers a wide range of neural voices but requires managing multiple projects and API keys. PressSynth streamlines this with a single API key and direct SDK compatibility.
Publishers can integrate PressSynth into their CMS or newsletter platforms with minimal setup. The straightforward pricing and lack of enterprise contracts make it easier to scale audio generation for daily news updates.
Unlike Google’s complex billing tiers, PressSynth’s per-character model ensures you only pay for what you generate, making it cost-effective for high-volume newsrooms.
- Single API key
- Simplified billing
- Direct SDK support
Pricing comparison
PressSynth charges $0.01 per 1,000 characters, with spaces and punctuation counting toward the total. Failed requests are free, ensuring you pay only for successful audio generation. This linear model scales predictably for newsrooms processing daily articles or newsletters.
Unlike competitors offering complex monthly tiers, PressSynth requires no minimum spend or seats. You top up via USDT on the TRON network, with balances that never expire. This structure suits publishers who need flexible, pay-as-you-go audio generation without long-term enterprise contracts.
Per-character billing model
The API bills at $0.01 per 1,000 characters. Every character, including spaces and punctuation, counts toward this limit. If synthesis fails, the charge is returned in full, so you only pay for usable audio files.
- $0.01 per 1,000 characters
- Spaces and punctuation count
- Failed requests are free
Monthly subscription alternatives
PressSynth does not offer monthly plans, seats, or minimum spend requirements. There are no subscriptions to manage or invoices with VAT to process. This simplicity allows newsrooms to adjust audio production volume based on daily editorial needs without commitment.
Free tier availability
New accounts signed in with Google receive $0.50 of credit, valid for 7 days. This covers approximately 50,000 characters, allowing teams to test voice quality and integration speed before committing to top-ups. Email-only accounts start with zero credit.
Technical integration
PressSynth integrates seamlessly into existing Python workflows by leveraging the official OpenAI SDK. Developers simply update the base_url to https://api.98wi.info/v1 and supply their unique API key. This approach eliminates the need for custom wrappers, allowing teams to reuse established TTS libraries with minimal configuration changes.
The API accepts standard JSON payloads, ensuring compatibility with common HTTP libraries. This design supports rapid prototyping and stable production environments alike, making it ideal for newsrooms automating article-to-audio pipelines without adopting proprietary SDKs.
OpenAI SDK compatibility
PressSynth mirrors the OpenAI speech endpoint structure, allowing direct use of the openai Python package. By overriding the base_url parameter, developers can switch providers without altering their core synthesis logic. This ensures consistent code behavior across different infrastructure setups.
- Compatible with official OpenAI Python SDK
- Simple
base_urlconfiguration override - Standard JSON request/response format
Rate limits per provider
Accounts are limited to 60 requests per minute per API key. New accounts start with a lower limit of 10 requests per minute until the first top-up is processed. This throttling ensures fair usage and maintains stability during high-volume news bursts.
Error handling capabilities
The API returns standard HTTP status codes for failures, such as 400 for invalid input or 429 for rate limit exceeded. Failed synthesis requests incur no charge, and credits are returned automatically. Developers should implement retry logic with exponential backoff to handle transient queue delays during peak hours.
Technical integration
PressSynth integrates seamlessly into existing Python workflows by leveraging the official OpenAI SDK. Developers simply update the base_url to https://api.98wi.info/v1 and supply their unique API key. This approach eliminates the need for custom wrappers, allowing teams to reuse established TTS libraries with minimal configuration changes.
The API accepts standard JSON payloads, ensuring compatibility with common HTTP libraries. This design supports rapid prototyping and stable production environments alike, making it ideal for newsrooms automating article-to-audio pipelines without adopting proprietary SDKs.
- Compatible with official OpenAI Python SDK
- Simple
base_urlconfiguration override - Standard JSON request/response format
Voice quality analysis
The eight preset voices deliver broadcast-ready clarity, leveraging licensed LibriTTS-R recordings. Each voice maintains consistent tone across the 23 supported languages, ensuring reliable output for diverse newsroom scripts. Generation speed approximates real-time, processing roughly 25 characters per second for standard texts.
Audio quality remains steady regardless of voice selection, from warm narrators to energetic promos. The system handles mixed-language content seamlessly, provided the language parameter is set for non-English Latin scripts. No emotion tags or SSML are available.
Naturalness of voices
Voices like Aria and Felix offer distinct characteristics suitable for different news formats. The synthesis engine produces clear, intelligible speech without excessive robotic artifacts. Speed adjustments between 0.25x and 4.0x allow fine-tuning without degrading audio integrity significantly.
- Aria: warm, clear narrator
- Felix: energetic, promo
- Atlas: neutral, news
- Orion: deep, documentary
Consistency across languages
The API supports 23 languages, with voices capable of speaking any of them. Script detection works for Latin text defaulting to English, but explicit language codes are recommended for accuracy. Non-supported languages like Vietnamese or Thai will not generate audio correctly.
Watermark presence
Every generated audio file contains an inaudible watermark identifying it as synthetic speech. This digital signature is embedded permanently within the file, distinguishing AI-generated news audio from human-recorded broadcasts without affecting listener experience.
Use case suitability
PressSynth’s eight preset voices map directly to common broadcast needs. Use atlas for neutral news delivery or luna for softer storytelling segments. The consistent engine ensures uniform audio quality whether you are generating daily headlines or weekend features.
The API handles variable text lengths well. Short alerts process quickly, while longer articles require splitting into 4096-character chunks. This approach keeps generation times predictable and avoids timeout issues for lengthy scripts.
News anchors vs commentary
The voice library offers distinct tones for different editorial styles. Atlas provides a neutral, news-ready delivery suitable for straight reporting. In contrast, luna’s soft, storytelling quality works better for commentary or human-interest pieces. Felix adds energy for promo content, while milo offers a conversational tone for interviews.
Each voice is distinct, allowing publishers to maintain brand identity across different content types. The consistent engine ensures that switching between voices does not alter the audio quality or background noise profile.
- Atlas: Neutral, ideal for straight news
- Luna: Soft, suitable for storytelling
- Felix: Energetic, good for promos
- Milo: Conversational, fits interviews
Short alerts vs long features
Short phrases typically take a few seconds to generate. For longer texts, the API requires splitting input into 4096-character segments. These segments are then joined into a single audio file. This method ensures consistent audio quality throughout long articles.
Generation speed is roughly real-time for long texts. Requests are processed sequentially, so high volume may introduce slight queues. This predictable timing helps publishers schedule audio releases for newsletters or broadcast slots.
- Max 4096 characters per request
- Sequential processing may cause queues
- Real-time speed for long texts
- Segments join into one file
Multi-language requirements
The API supports 23 languages, covering major global markets. Any voice can speak any supported language, providing flexibility in voice selection. For non-English Latin text, explicitly setting the language parameter prevents incorrect guessing.
Unsupported languages include Vietnamese and Thai. Publishers targeting these regions must rely on English narration or manual translation. The clear support list helps avoid unexpected language errors in automated workflows.
- 23 supported languages
- Any voice speaks any supported language
- Explicit language setting recommended
- Vietnamese and Thai not supported

Data privacy comparison
PressSynth prioritizes input confidentiality by using text solely for audio synthesis, ensuring it is not repurposed for model training. This approach is particularly valuable for newsrooms handling sensitive drafts or embargoed stories before publication. The generated audio files are returned directly in the API response and are not retained on our servers, providing an ephemeral workflow suitable for time-sensitive digital newsletters.
Training data usage policies
Unlike some competitors that ingest user data to improve their base models, PressSynth keeps input text isolated. Your article content remains yours, used only to generate the immediate audio file. This distinction eliminates the risk of your unpublished news copy influencing future public voice outputs.
- Input text is not used for training.
Audio storage practices
Generated audio is transient. Once the API returns the file to your client or application, no copy is stored on PressSynth servers. This reduces data residency complexity for publishers who need to manage their own audio archives independently. Only request metadata, such as timestamp and character count, is retained for billing verification.
Enterprise contract options
PressSynth operates on a transparent, self-serve model rather than offering bespoke enterprise agreements. There are no custom data processing addendums or dedicated support tiers. While this simplicity suits most digital newsrooms, it means there is no formal SLA or on-premise installation option for organizations requiring strict contractual data guarantees.
- No SLA or uptime guarantee.
Support and SLA
PressSynth operates without an enterprise contract or phone number, relying entirely on email for support. There is no live chat option available. This approach keeps overhead low, which helps maintain the simple per-character pricing model.
The service does not offer an SLA or uptime guarantee, meaning availability is not contractually bound. While requests are processed sequentially, queues may form under load. Publishers should account for this variability when scheduling automated newsletter audio generation to avoid tight deadlines.
Email vs phone support
All inquiries are handled via email, with no dedicated phone line or live chat widget. This means response times may vary depending on ticket volume. For newsrooms, this asynchronous model works well for routine API questions but requires patience for urgent integration issues.
Support is limited to email only, so you cannot call for immediate help. Ensure your team has a designated contact person to monitor support emails regularly. This structure supports the product’s lean operational model, avoiding the costs associated with 24/7 call centers.
- Email-only support channel
- No dedicated phone number
- No live chat widget
- Asynchronous response model
Uptime guarantees
PressSynth does not provide a formal SLA or uptime guarantee. There is no enterprise contract to back up service level promises. Requests are processed one after another, so high traffic can lead to queuing delays. This lack of guaranteed availability is reflected in the lower price point.
Since there is no on-premise install or data-center choice, service depends entirely on the cloud provider’s stability. Publishers should design their workflows to handle occasional delays gracefully. For critical, time-sensitive broadcasts, consider generating audio files in advance rather than relying on real-time generation.
- No formal SLA
- No uptime guarantee
- No enterprise contract
- Requests processed sequentially
Documentation quality
The documentation is clear and practical, focusing on the OpenAI SDK compatibility. It includes ready-to-use curl examples and explains the request shape precisely. This makes it easy for developers to integrate the API into Python scripts without guessing endpoints or headers.
Support pages cover rate limits, voice options, and error handling. The guides are straightforward, avoiding unnecessary jargon. However, advanced features like custom pronunciation are not available, so the docs accurately reflect the product’s current capabilities without over-promising.
- Clear curl examples
- OpenAI SDK compatibility explained
- Rate limits clearly stated
- Voice options listed
Python ecosystem fit
PressSynth does not offer an official proprietary SDK. Instead, it relies on the OpenAI SDK ecosystem, requiring developers to adjust the base_url to https://api.98wi.info/v1. This approach ensures immediate compatibility with existing Python libraries and simplifies integration for teams already familiar with standard TTS workflows.
The Python community benefits from this standardized endpoint structure. Developers can utilize familiar HTTP clients or the official OpenAI Python package without learning a new API syntax. This reduces the learning curve for newsrooms adopting automated audio generation for newsletters and digital broadcasts.
Official SDK availability
There is no dedicated PressSynth Python library. Integration uses the standard OpenAI SDK by overriding the base URL. This means your existing code structures remain valid, requiring only a configuration change to point to the PressSynth API endpoint.
- Uses standard OpenAI Python package
- No proprietary client library needed
- Simple base_url configuration
Community support
Because the API mirrors the OpenAI format, existing community resources for Python TTS integration are largely applicable. Developers can find ample examples of handling JSON responses and audio streams in Python forums, reducing the need for niche troubleshooting.
Documentation clarity
The /docs/ page provides clear Python-friendly examples. It details the required headers, JSON body structure, and error codes. The documentation is straightforward, focusing on practical implementation steps for developers integrating audio generation into their publishing pipelines.
Final recommendation
PressSynth delivers reliable, scalable audio generation for publishers who need clear, broadcast-quality output without enterprise contracts. Its straightforward pricing and broad language support make it a practical choice for digital newsrooms.
The API’s compatibility with standard OpenAI SDKs simplifies integration, allowing teams to convert articles and scripts into audio efficiently. This eliminates the need for custom infrastructure.
Best for budget-conscious newsrooms
At $0.01 per 1,000 characters, PressSynth offers predictable costs without monthly subscriptions or minimum spend. Failed requests are free, ensuring you only pay for successful synthesis.
The absence of enterprise contracts and VAT invoices simplifies accounting. Accounts start with a small free trial, making it easy to test volume before committing funds via USDT.
- No monthly plans or seats
- Pay only for successful requests
- Simple USDT top-ups
Best for simple Python integration
The API mirrors the OpenAI speech endpoint, allowing direct use of official OpenAI SDKs by changing the base URL. This means existing Python code for TTS can often be adapted with minimal changes.
Rate limits are clearly defined at 60 requests per minute, with queueing during high load. This predictability helps newsrooms plan automated article-to-audio workflows without unexpected throttling.
- OpenAI SDK compatible
- Straightforward JSON requests
- Clear rate limits
Best for multi-language needs
Support for 23 languages, including Arabic, Chinese, Hindi, and Swahili, allows global news coverage without managing separate voice providers. Any voice can speak any supported language, simplifying voice selection.
Automatic language detection works for non-English Latin text if the language parameter is set. This reduces errors and ensures accurate pronunciation across diverse linguistic content.
- 23 languages supported
- Any voice, any language
- Automatic language detection
How to get started
Getting started with PressSynth takes minutes. Visit /get-api-key/ to sign in with Google or create an email account. You receive $0.50 in trial credit immediately, valid for seven days, allowing you to generate roughly 50,000 characters of audio before paying.
Once registered, your API key appears on the dashboard. Use this key to authenticate requests. You can top up via USDT on the TRON network when needed, with automatic crediting within minutes.
Sign up process
Navigate to /get-api-key/ and choose Google Sign-In or email registration. Accounts are specific to 98wi.info. No credit card is required to begin testing the service.
API key generation
Your active key displays on the dashboard alongside your balance and usage stats. Only one key is active at a time; issuing a new one instantly invalidates the previous one.
First request example
Send a POST request to https://api.98wi.info/v1/audio/speech. Use the standard OpenAI SDK structure. Set the base URL to https://api.98wi.info/v1. Include your model, input text, and desired voice. The response returns the audio file directly.
Example: `curl -X POST https://api.98wi.info/v1/audio/speech -H "Authorization: Bearer YOUR_KEY" -d '{"model": "tts-1", "input": "Breaking news update.", "voice": "aria"}'`.
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.
23 languages
Any voice speaks any of these languages.
