Frequently asked questions
Which voice is best for news?
Q: Which voice is suitable for news? A: Atlas is designed as a neutral, news-style voice. Orion offers a deep, documentary tone suitable for serious reporting. Felix provides an energetic, promo-style delivery for breaking updates. Nova is bright and upbeat for lighter segments. All eight voices support all 23 languages, allowing editorial teams to match the voice to the specific tone of each news script.
How does pricing compare to competitors?
NewsVoice API charges $10 per 1 million characters, or $0.01 per 1,000 characters. This includes spaces and punctuation. There are no monthly plans, seats, or minimum spends. Failed requests are free, with charges returned in full. This flat-rate model allows editorial teams to predict costs precisely based on character volume without tiered pricing complexities.
Is Google Docs read aloud better?
Google Docs read-aloud is a manual, browser-based feature. Our API provides programmatic access for automated, high-volume generation suitable for broadcast pipelines. You can integrate synthetic voices directly into your workflow via JSON requests. Unlike manual playback, our API supports multiple formats (mp3, wav, opus) and allows precise control over speed and voice selection for digital distribution.
What languages are best supported?
We support 23 languages: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. Any voice can speak any of these languages. If the language is not specified, it is guessed from the script, defaulting to English for Latin text.
Can I clone a news anchor's voice?
No. We do not offer custom voice cloning or uploading of your own voice. Voices are based on licensed recordings from LibriTTS-R and LibriVox. Each generated file carries an inaudible watermark marking it as synthetic speech. You must choose from the eight available presets (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion) for your news scripts.
Is there a free trial?
Yes. Signing in with Google provides $0.50 of credit, valid for 7 days, covering approximately 50,000 characters. Accounts created with email and password start without trial credit. There are no monthly plans or subscriptions. The balance does not expire, and you can top up with USDT on the TRON network at any time.
How fast is the generation?
Generation is roughly real-time on long texts, processing about 25 characters per second. Short phrases may take a few seconds. Requests are processed sequentially, so under load, a request may wait in a queue. For longer texts, split input into parts of up to 4,096 characters and join the resulting audio files programmatically.
Do I need technical skills?
Yes, basic technical skills are required. You need to send a POST request to the API endpoint with a JSON body. The request shape matches the OpenAI speech endpoint, so official OpenAI SDKs work by changing the base URL. You must handle authentication with a Bearer token and manage character limits and audio format selection in your code.
Is data private?
Input text is used only to produce the audio and is not used for training. Generated audio is returned in the response and not stored. We keep request metadata (time, characters, status) for billing purposes. We do not offer an on-premise install or specific data-center choices. There is no GDPR certificate, but data usage is limited.
What happens if requests fail?
If synthesis fails, the charge is returned in full. Failed requests are free. This ensures you only pay for successfully generated audio files. You can monitor the status in your account dashboard. The API returns an error response if the request cannot be processed, allowing your application to retry or handle the failure gracefully without cost.
Can I use it for live news?
The API processes requests sequentially, which may cause queuing under load. Generation is roughly real-time, but short phrases take a few seconds. It is not a streaming WebSocket service. For live news, you should pre-generate audio files or split long texts into smaller parts. The 60 requests per minute limit may constrain high-frequency live updates.
Do you offer an SLA?
No. We do not offer an SLA, uptime guarantee, or enterprise contract. There is no phone number or live chat for support; assistance is by email only. We do not provide certifications like SOC 2 or HIPAA. The service is designed for editorial teams who can tolerate variable latency and lack formal uptime commitments.
How do I pay with USDT?
Top-ups are accepted via USDT on the TRON network (TRC-20) only. The minimum top-up amount is $5. Credits are added automatically within minutes after the transfer. There are no cards, PayPal, or invoices with VAT. You can issue a new API key at any time, which stops the old one immediately, ensuring secure payment management.
Is there a mobile app?
No. We do not offer a mobile app, browser extension, or WordPress plugin. You interact with the API via HTTP requests or compatible SDKs like the OpenAI SDK. There is no official SDK from us, no speech-to-text, and no translation services. You must build your own interface or integrate the API into your existing editorial tools.
Key Criteria for Selection
For newsrooms generating audio at scale, the API’s volume capacity hinges on processing 4096-character chunks sequentially. With a throughput of roughly 25 characters per second, long scripts generate audio at roughly real-time speed, though requests queue under heavy load. Teams must split longer texts to maintain steady output without exceeding the 60 requests per minute limit per key.
Volume capacity for high traffic
High-volume news desks should implement client-side text splitting to manage the 4096-character limit per request. Since requests process sequentially, bursts of traffic may experience queue delays. Monitoring the rate limit is essential for maintaining consistent audio delivery during breaking news cycles.
- Process up to 4096 characters per request
- Sequential processing creates queues under load
- 60 requests per minute per API key
Language support for global reach
The API supports 23 languages, covering major global news markets including English, Spanish, German, French, and Mandarin. Any voice can speak any supported language, allowing consistent audio branding across regions. The system auto-detects language from script, but explicit language codes are recommended for Latin-script texts outside English to ensure accuracy.
- 23 supported languages including Arabic, Japanese, and Swahili
- Any voice works with any supported language
- Auto-detection defaults to English for Latin script
Cost efficiency for large teams
Pricing is strictly usage-based at $10 per million characters, eliminating fixed monthly fees or seat costs. This model suits large teams with variable output needs, as you only pay for generated audio. Failed requests incur no charge, protecting the budget against technical errors during high-volume deployments.
- $0.01 per 1,000 characters
- No monthly plans or seat restrictions
- Failed requests are free
Voice Options Compared
Selecting the right voice model depends on the specific tone of your broadcast. News teams typically require distinct auditory profiles for different content types, from hard news to feature stories.
The API offers eight presets. Each voice is derived from licensed recordings, ensuring a consistent and professional sound across your distribution channels.
Atlas for neutral news reporting
Atlas provides a neutral tone ideal for standard news delivery. Its clear, unobtrusive style ensures the focus remains on the script's facts rather than the narrator's personality.
This preset is well-suited for breaking news updates and wire service reads where neutrality is critical. It avoids emotional inflection that might bias the audience.
- Neutral tone
- Ideal for hard news
- Unobtrusive delivery
Orion for documentary style
Orion offers a deep, documentary-style voice. Its rich timbre adds gravitas to feature stories, long-form investigations, and contextual background pieces.
Use Orion when you need to establish authority and depth. The voice carries a narrative weight that engages listeners over extended durations without fatigue.
- Deep, authoritative tone
- Documentary and feature stories
- High gravitas
Sage for instructional content
Sage delivers a calm, instructional tone perfect for educational segments or procedural updates. It is clear and steady, reducing listener cognitive load.
Ideal for explainers, weather briefings, or instructional news segments. Sage’s pacing and clarity help ensure complex information is absorbed accurately by the audience.
- Calm and instructional
- Clear pacing
- Ideal for explainers
Language Support Analysis
NewsVoice API supports 23 languages, covering the vast majority of global news markets. This includes major languages like English, Spanish, French, German, and Japanese, ensuring your broadcasts reach diverse audiences without needing separate voice providers for each region.
The system handles non-Latin scripts, including Arabic, Chinese, Hebrew, Hindi, Korean, Russian, and Turkish. Smaller language support like Swahili, Malay, and Greek is also available, allowing for comprehensive multilingual content strategies in a single integration.
23 languages supported
The API covers 23 specific languages, including Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish.
This selection ensures that most major international news desks can generate audio content directly without third-party localization services. Note that languages like Vietnamese, Indonesian, Thai, Ukrainian, Romanian, and Czech are not currently supported.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
Major news languages included
Core languages such as English, Spanish, French, German, and Japanese are fully supported with all eight voice presets. Each voice can speak any of the 23 languages, allowing you to maintain consistent audio branding across different linguistic regions.
This flexibility means you can assign specific voices like Atlas for neutral news reporting in English or Spanish, ensuring a uniform tone regardless of the language used in the script.
Automatic detection features
If the language parameter is omitted, the API guesses the language based on the script. Latin-script text defaults to English, so you must explicitly set the language parameter for non-English Latin text (e.g., French or German) to ensure correct pronunciation.
For scripts using non-Latin scripts like Arabic or Chinese, the system automatically detects the language, eliminating the need for manual configuration in most cases.
Pricing Models Reviewed
NewsVoice API uses a straightforward per-character pricing model, eliminating the need for monthly subscriptions or seat-based fees. At $10 per million characters, costs scale directly with usage, making it ideal for large editorial teams with variable output needs. This model ensures you pay only for what you generate, with no minimum spend.
Failed requests are refunded in full, ensuring you are never charged for unsuccessful synthesis attempts. The absence of monthly plans or complex tiered structures simplifies budgeting for newsrooms, allowing finance teams to predict costs based on actual character volume rather than abstract usage caps.
Per character vs monthly plans
The API charges $0.01 per 1,000 characters, including spaces and punctuation. This transparent rate applies uniformly regardless of voice or model choice. Unlike traditional SaaS tools that lock you into annual contracts, this model offers flexibility for teams experiencing fluctuating news cycles.
There are no monthly plans, seats, or minimum spend requirements. This structure supports large teams where not every member generates audio daily, avoiding waste on unused licenses. Balance does not expire, allowing teams to top up during high-volume periods and draw down during quieter weeks.
- $10 per 1 million characters
- $0.01 per 1,000 characters
- No monthly subscriptions
- No seat limits
- No minimum spend
Free trial availability
New accounts created via Google sign-in receive a $0.50 trial credit, equivalent to approximately 50,000 characters, valid for 7 days. This allows editorial teams to test voice quality and integration workflows without immediate financial commitment. Accounts created with email and password do not receive this trial credit.
This trial period is sufficient for most news teams to evaluate the API's performance with typical news script lengths before committing to paid usage.
- $0.50 credit for Google sign-ups
- Approximately 50,000 characters
- Valid for 7 days
- Email/password accounts start without credit
USDT payment option
Top-ups are processed via USDT on the TRON network (TRC-20), with a minimum deposit of $5. Credits are automatically added within minutes of transfer confirmation, ensuring minimal downtime for urgent news cycles. This method avoids credit card processing fees and supports global teams without regional banking restrictions.
There are no invoices, VAT charges, or recurring billing cycles. Each top-up is a one-time transaction, giving teams full control over their cash flow and expenditure timing.
- USDT (TRC-20) only
- Minimum $5 top-up
- Automatic credit within minutes
- No VAT or invoices
- No credit cards or PayPal
Technical Integration Ease
Integration requires minimal overhead for newsrooms already familiar with standard REST patterns. The API mirrors the OpenAI speech endpoint structure, allowing teams to reuse existing libraries or adapt simple scripts without learning a proprietary protocol. This compatibility significantly reduces onboarding time for engineering teams managing high-volume broadcast workflows.
Developers can switch from other providers by simply updating the base URL and injecting their unique API key. The consistent request shape ensures that existing error-handling logic and retry mechanisms remain valid, streamlining the transition to a dedicated news-focused TTS solution.
OpenAI SDK compatibility
The API accepts standard JSON payloads, meaning you can use the official OpenAI SDKs by pointing them to https://api.playwin333.com/v1. This allows you to leverage familiar client libraries for Python, Node.js, or Go, reducing the need for custom HTTP wrappers.
Authentication is handled via a single Bearer token in the header. This key can be regenerated instantly if compromised, ensuring secure access for multiple team members without complex permission structures.
- Works with official OpenAI SDKs by changing
base_url - Standard
Bearertoken authentication - No proprietary client libraries required
Simple cURL examples
You can test the endpoint immediately with a single cURL command. Sending a POST request to /v1/audio/speech with your JSON body returns the audio file in the response.
This approach is ideal for quick validation of voice settings, speed adjustments, or language detection before integrating into larger editorial pipelines or automation scripts.
- Direct JSON body submission
- Immediate audio file return
- Ideal for quick validation
Standard JSON API
The request structure uses standard fields like model, input, voice, and response_format. This simplicity means you can easily parse the response in any language, whether you are building a custom broadcast automation tool.
Errors return standard HTTP status codes, making it easy to integrate with existing monitoring tools and alert systems used by large editorial teams.
- Standard
model,input,voicefields - Returns standard HTTP status codes
- Easy parsing in any language
Scalability for Large Teams
For high-volume newsrooms, understanding rate limits is critical to maintaining consistent audio generation pipelines. Each API key supports 60 requests per minute, allowing editorial teams to process multiple news scripts simultaneously without bottlenecking the workflow. Before any top-up, the limit is lower at 10 requests per minute, which suits initial testing but requires upgrading for daily broadcast needs.
The system processes requests sequentially, meaning that during peak traffic, a request may wait in a queue before synthesis begins. This ensures fair resource distribution but requires teams to plan their generation schedules accordingly, especially when producing large batches of audio for digital distribution.
Rate limits per key
Each API key supports 60 requests per minute, enabling consistent throughput for daily news updates. New accounts start with 10 requests per minute, which scales up after the first top-up. This predictable limit allows engineering teams to configure their automation scripts with appropriate retry logic and backoff strategies.
- 60 requests/minute per key after top-up
- 10 requests/minute for new accounts
- Predictable throughput for automation
Queue processing under load
Requests are processed one after another, so under heavy load, a request may wait in a queue. This sequential processing ensures stability but means generation times can vary during peak hours. Teams should account for this latency when scheduling real-time updates.
- Sequential processing ensures stability
- Latency may increase under load
- Plan for queue delays
No seat restrictions
There are no limits on the number of team members or seats. All editors and developers can share the same API key or use individual keys, depending on your security preferences. This flexibility supports large editorial teams without additional per-user costs.
- Unlimited team members
- No per-seat licensing fees
- Flexible key management
Audio Quality Assessment
NewsVoice API generates audio from licensed recordings, ensuring a consistent and professional baseline for broadcast quality. The voices are derived from LibriTTS-R and LibriVox public-domain readings, providing a reliable foundation for news scripts and updates.
The API uses a TTS model that prioritizes clarity and neutrality. The output is designed for functional, clear communication.
Licensed recording base
The voice models are trained on licensed recordings from LibriTTS-R and public-domain audiobook readings from LibriVox. This ensures that the synthetic voices have a natural, human-like quality without the uncanny valley effect often associated with basic TTS systems.
Each voice preset—whether neutral like Atlas or deep like Orion—maintains a consistent tone throughout long texts. This consistency is critical for news teams producing extended broadcasts or series of audio updates where listener fatigue can set in.
- Voices based on LibriTTS-R (CC BY 4.0) recordings
- Derived from LibriVox public-domain audiobook readings
- Consistent tone across long-form content
Synthetic watermark presence
Every audio file generated by the API carries an inaudible watermark. This watermark marks the content as synthetic speech, which is useful for copyright tracking and distinguishing AI-generated news from human-recorded broadcasts.
The watermark is embedded during synthesis and does not affect the audible quality or clarity of the speech. It remains in the final output files regardless of the selected format (mp3, wav, opus, flac, or pcm).
- Inaudible watermark in every generated file
- Marks audio as synthetic speech
- Does not affect audible quality
Natural speech flow
The API processes requests sequentially, which can lead to queueing under heavy load. However, the output speech flows naturally, with proper pacing and intonation that mimics human news delivery.
Speed can be adjusted between 0.25 and 4.0 to match the desired broadcast tempo. For standard news reading, a speed of 1.0 provides a clear, professional cadence suitable for digital distribution and radio simulcasts.
- Adjustable speed from 0.25 to 4.0
- Natural pacing and intonation
- Suitable for standard news reading at 1.0x speed
Data Privacy Comparison
News teams prioritize data integrity when broadcasting sensitive scripts. Our API ensures that the input text used for synthesis is strictly processed to generate the audio file and is not retained for training future models. This separation guarantees that your editorial content remains proprietary and does not influence the underlying voice engine.
Input text not trained on
The text you submit is used solely for the immediate production of the audio file. It is not ingested into any training dataset, ensuring that your unique news scripts and editorial voice remain distinct from the model’s core data.
Audio not stored
Generated audio files are returned directly in the API response and are not stored on our servers after delivery. This ephemeral handling means your broadcast assets remain under your control, with no persistent copies held by the provider.
Minimal metadata retention
We retain only the request metadata necessary for accurate billing and service monitoring. This includes the timestamp, character count, and request status. No content or audio files are kept beyond what is required for operational accounting.
Support and Documentation
NewsVoice API provides documentation and support tailored for technical teams managing high-volume news production. Our documentation is accessible at /docs/ and covers all endpoints, including voice selection and format options. We prioritize clarity to ensure your engineers can integrate the API quickly without unnecessary overhead.
For direct assistance, support is available exclusively via email. This approach ensures that every query receives a thorough, documented response without the noise of live chat or the delays of phone queues. You can reach us at /contact/ for any integration or billing questions.
Email support only
Our support team handles all inquiries through email. This allows for detailed, written responses that can be easily referenced by your entire team. Unlike phone support, email ensures that complex technical questions about rate limits or voice presets are answered with precision.
There is no live chat or phone number to call. This structure is designed for efficiency, allowing support engineers to focus on resolving your specific integration issues without interruption. You can submit your request via our contact page for timely assistance.
- Direct email contact for all inquiries
- Written responses for easy team reference
- No live chat or phone support
Comprehensive API docs
Our documentation provides clear examples for every endpoint, including the speech generation endpoint and voice listing. You will find ready-to-use cURL commands and explanations for the JSON body structure, making it easy to start generating audio immediately. The docs also cover authentication and rate limits.
We include details on all 8 voice presets and 23 supported languages, helping you choose the right tone for your news content. The documentation is updated regularly to reflect any changes in the API, ensuring your team always has access to the latest information.
- Ready-to-use cURL examples
- Details on all 8 voice presets
- Authentication and rate limit guides
Clear error messages
When a request fails, you will receive a clear error message indicating the issue. This could be related to character limits, invalid voice IDs, or authentication problems. Our error messages are designed to help you debug your integration quickly, minimizing downtime for your news broadcasts.
Each error response includes a status code and a descriptive message, allowing your team to handle failures programmatically. This clarity ensures that your automated systems can retry or alert your team effectively, maintaining the reliability of your audio distribution pipeline.
- Status codes and descriptive messages
- Helps debug character limit issues
- Enables programmatic failure handling
Limitations to Consider
News teams requiring real-time audio injection into live broadcasts should note the absence of streaming capabilities. The API returns complete audio files after generation, meaning partial audio chunks are not available via WebSocket during synthesis.
This architecture suits scheduled updates and on-demand clips but may introduce slight latency for live breaking news scenarios where immediate audio drop-in is critical.
No streaming or WebSocket
The standard endpoint delivers finished audio files rather than a continuous stream. This means you cannot receive audio data in real-time chunks, which is necessary for some live broadcast automation systems.
No SSML support
The API does not support Speech Synthesis Markup Language (SSML). You cannot use SSML tags to control pronunciation, pause duration, or pitch adjustments within the JSON body.
- Basic speed control is available via the
speedparameter (0.25–4.0). - Voice selection is limited to the eight predefined presets.
No custom voice cloning
Customers cannot upload audio samples to create custom voice clones. The service relies strictly on the eight licensed presets derived from LibriTTS-R and LibriVox recordings.
Use Case Recommendations
News teams leverage the API to convert breaking scripts into immediate audio updates. By splitting texts exceeding 4096 characters, editorial desks ensure continuous playback without manual intervention, maintaining pace with developing stories.
For global reach, the system supports 23 languages, allowing a single voice preset to deliver localized content across markets. Automatic language detection simplifies workflows for diverse editorial teams managing international wire feeds.
News websites and apps can use the API to provide audio versions of articles. This direct programmatic access eliminates third-party dependencies, ensuring reliable delivery for readers requiring accessible content formats without extra infrastructure.
Breaking news audio updates
Editorial desks can send scripts directly to the API, receiving audio files quickly. With a typical speed of roughly real-time, long breaking news updates generate efficiently, allowing teams to publish audio alongside text.
The system handles texts up to 4096 characters per request. For longer scripts, splitting and joining audio files ensures seamless playback, keeping listeners engaged during rapid updates without manual editing.
- Convert breaking scripts to audio instantly
- Split long texts for continuous playback
- Maintain pace with developing stories
Multilingual broadcast content
The API supports 23 languages, including major news languages like English, Spanish, and French. Any voice preset can speak any supported language, simplifying voice selection for global content strategies.
Automatic language detection guesses the language from the script, but setting the language field explicitly ensures accuracy for non-English Latin-script text, preventing mispronunciations in critical broadcasts.
- Support for 23 languages
- Any voice can speak any supported language
- Explicit language setting for accuracy
Accessibility compliance
News websites and apps can use the API to provide audio versions of articles. This direct programmatic access eliminates third-party dependencies, ensuring reliable delivery for readers requiring accessible content formats without extra infrastructure.
Generated audio files are returned in the response and not stored, keeping data minimal. This approach supports accessibility standards while maintaining privacy, as input text is used only for synthesis.
- Provide audio versions of articles
- Direct programmatic access
- Eliminate third-party dependencies
Final Recommendation Summary
This API is suitable for high-volume newsrooms requiring reliable, programmatic audio generation for digital distribution and broadcast updates. It handles large character counts efficiently, processing requests sequentially with predictable timing for long scripts.
The pricing model eliminates monthly seats and minimum spend, charging only for characters used. This structure ensures large teams pay strictly for what they generate, with no hidden fees or subscription lock-ins.
Integration is straightforward via the standard OpenAI SDK or plain HTTP. Changing the base URL allows immediate use of existing codebases, while comprehensive documentation and clear error messages reduce onboarding time for engineering teams.
Best for high-volume newsrooms
This API is suitable for high-volume newsrooms requiring reliable, programmatic audio generation for digital distribution and broadcast updates. It handles large character counts efficiently, processing requests sequentially with predictable timing for long scripts.
- Processes up to 4096 characters per request
- Sequential processing ensures predictable queue times
- Ideal for breaking news audio updates
Cost-effective for large teams
The pricing model eliminates monthly seats and minimum spend, charging only for characters used. This structure ensures large teams pay strictly for what they generate, with no hidden fees or subscription lock-ins.
- $0.01 per 1,000 characters
- No monthly plans or seat limits
- Failed requests are free
Easy integration with existing tools
Integration is straightforward via the standard OpenAI SDK or plain HTTP. Changing the base URL allows immediate use of existing codebases, while comprehensive documentation and clear error messages reduce onboarding time for engineering teams.
- Compatible with official OpenAI SDKs
- Standard JSON API structure
- Simple cURL examples provided
Listen to the 8 voices
Each sample below was generated by this API from the same sentence.
23 languages
Any voice speaks any of these languages.
Sources and further reading
- Wikipedia
- openai-python on GitHub
- W3C WAI: making audio and video accessible
- TRON developer hub: TRC-20
- OpenAI API reference: create speech
- LibriTTS-R corpus (OpenSLR 141)
- IETF RFC 7845: Ogg Opus
- Node.js documentation
- curl documentation
- Python documentation
- FFmpeg documentation
- LibriTTS-R paper (arXiv 2305.18802)