Frequently asked questions
What is a TTS API?
A cloud text-to-speech API that accepts text and returns an audio file via a POST request. It uses standard JSON bodies compatible with OpenAI SDKs, allowing news teams to programmatically convert scripts into synthetic speech files for digital distribution without manual recording.
How does it differ from Google Docs read aloud?
Unlike Google Docs read aloud, which is a browser-based listening feature, this API provides programmatic access to synthetic voices. It generates audio files (MP3, WAV, etc.) suitable for broadcast and digital distribution, rather than just reading text aloud in a document interface.
Is it suitable for large news teams?
Yes. Designed for high-volume generation, it supports 60 requests per minute per key. Teams can split longer texts into 4096-character parts and join the audio files, ensuring reliable delivery for news scripts and updates across large editorial workflows.
What is the cost per character?
Pricing is $10 per 1 million characters, or $0.01 per 1,000 characters. Spaces and punctuation count as characters. Failed requests are free, with charges returned in full if synthesis fails, ensuring you only pay for successful audio generation.
Which languages are supported?
Supports 23 languages including English, Spanish, French, German, Arabic, Chinese, Japanese, and Hindi. Any voice can speak any supported language. If the language is not specified, it is guessed from the script, defaulting to English for Latin text.
Can I use it for breaking news?
Yes. Texts are processed sequentially, taking roughly real time (25 characters per second). Short phrases generate in seconds, allowing rapid production of audio updates. You can split longer scripts into parts to manage queue times during high-volume breaking news events.
Do I need technical skills?
Basic technical knowledge is needed to use the API endpoint. You send a JSON body with text, voice, and format parameters. The interface matches the OpenAI speech endpoint, so existing OpenAI SDKs work by simply changing the base URL to our API.
How is audio generated?
Audio is generated by sending text to the POST /v1/audio/speech endpoint. The system synthesizes the input using selected voices like Aria or Felix. The resulting audio file is returned in the response, ready for download or immediate playback in your workflow.
Is the voice realistic?
Voices are based on licensed recordings from LibriTTS-R and LibriVox public-domain audiobooks. Each generated file carries an inaudible watermark marking it as synthetic speech. The voices offer clear, natural-sounding narration suitable for news reading, though they are distinct synthetic voices.
Can I clone a specific voice?
No. Custom voice cloning is not offered. You choose from eight preset voices (e.g., Aria, Felix, Atlas) based on licensed recordings. You cannot upload your own voice or design a custom voice profile for specific news anchors or personalities.
Is there a free trial?
Yes. Signing in with Google provides $0.50 of credit (about 50,000 characters) valid for 7 days. Accounts created with email and password start without trial credit. This allows teams to test the API’s quality and integration before purchasing credits.
What payment methods are accepted?
We accept USDT on the TRON network (TRC-20) only. There are no credit cards, PayPal, or invoices. Top-ups are minimum $5 and credited automatically within minutes. Balance does not expire, and there are no monthly plans or subscriptions required.
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. Request metadata (time, characters, status) is kept solely for billing purposes, ensuring data privacy for news scripts.
Do you offer an SLA?
No. We do not offer an SLA, uptime guarantee, or enterprise contract. Requests are processed sequentially, so under load, a request may wait in a queue. Support is by email only, with no phone number or live chat available.
Definition of TTS API
A TTS API is a programmatic text-to-speech service designed for automated audio generation. It functions as a cloud-based infrastructure endpoint where newsrooms can send text data and receive synthesized audio files programmatically, eliminating manual recording steps.
This system enables editorial teams to convert scripts into audio instantly via standard HTTP requests. The cloud architecture ensures reliable scalability for high-volume news production without requiring on-premise hardware management.
Programmatic text-to-speech service
The NewsVoice API provides a direct interface for software to generate speech. Editorial workflows can trigger audio creation directly from content management systems or scripts, ensuring consistent voice output across digital and broadcast channels.
Unlike static audio files, this service allows dynamic generation of news updates. Teams can update the script and regenerate the audio instantly, maintaining accuracy for breaking news without re-recording studio sessions.
- Supports JSON payloads similar to OpenAI's speech endpoint
- Uses standard HTTP methods for easy integration
Send text, receive audio file
The core interaction involves submitting text input and receiving an audio file. Users send a POST request to the API endpoint with the script text, specifying parameters like voice, language, and format. The system processes the request and returns the synthesized audio.
This method supports up to 4096 characters per request. For longer news scripts, texts are split into parts, synthesized individually, and joined. This ensures high-quality audio delivery regardless of script length.
Cloud-based infrastructure
The service operates entirely in the cloud, requiring no local installation or heavy hardware. Newsrooms can scale audio production based on demand, handling peak loads during major events without infrastructure bottlenecks.
Requests are processed sequentially, ensuring consistent quality. The cloud environment handles the computational load of speech synthesis, allowing editorial teams to focus on content creation rather than technical maintenance.
How It Works for Newsrooms
The API transforms editorial workflows by accepting news scripts via HTTP and returning audio files, eliminating manual recording delays. Newsrooms can convert text to speech instantly, ensuring updates are available the moment a story is filed.
Convert scripts to audio instantly
Send JSON payloads with your script to the speech endpoint. The system returns an audio file in your chosen format within seconds, enabling rapid distribution of breaking news without waiting for studio availability.
Automate voiceover production
Integrate the API with your content management system to trigger audio generation automatically upon publication. This removes the bottleneck of human voiceover artists for routine updates, allowing editorial teams to scale output efficiently.
- Trigger audio on publish via webhook or CMS plugin
- Reduce manual voiceover scheduling overhead
- Maintain consistent delivery speed during high-volume news cycles
Scale across multiple languages
Generate multilingual broadcasts from a single script source. The API supports 23 languages, allowing news outlets to serve diverse audiences without hiring separate voice talent for each market, ensuring consistent brand tone across regions.
Difference from Google Docs Read Aloud
Google Docs read-aloud is a consumer-facing feature designed for individual users reading documents on screen. It lacks programmatic access, meaning you cannot trigger audio generation via code or integrate it directly into your newsroom’s content management system. You must manually open the document and play the audio, which does not scale for large editorial teams generating multiple updates daily.
In contrast, a TTS API like NewsVoice allows you to send text via a simple HTTP request and receive an audio file instantly. This enables automation: when a journalist publishes a script in your CMS, the API can immediately generate the audio file and push it to your broadcast pipeline or website without human intervention.
API allows programmatic control
You control the entire process via standard HTTP requests. Send a JSON payload with your script, and receive the audio file directly. This eliminates manual steps and integrates seamlessly with existing editorial workflows.
Unlike Google Docs, which requires user interaction, the API processes requests in the cloud. Your team can generate audio for breaking news updates programmatically, ensuring consistent delivery speeds regardless of how many editors are working.
- Send text via JSON; receive audio file
- Trigger generation on content publish
- No manual UI interaction required
Higher volume capacity
Google Docs is limited by browser resources and user sessions. A TTS API processes requests sequentially, supporting the needs of large newsrooms generating dozens or hundreds of audio files per day.
With a rate limit of 60 requests per minute per key, you can scale your production. Short phrases generate in seconds, while longer news scripts process in near real-time, ensuring timely distribution of updates.
Dedicated editorial voices
Google Docs uses generic, robotic voices that vary by browser and OS. NewsVoice provides 8 consistent, licensed presets like Atlas (neutral, news) or Felix (energetic, promo), ensuring a uniform brand voice across all broadcasts.
These voices are based on licensed recordings and include an inaudible watermark to mark content as synthetic. You can select the exact tone that matches your station’s identity, rather than accepting whatever the browser provides.
Integration Workflow
Connect the API to your CMS. When a story is marked 'ready for audio', your system sends the text to the API. The API returns the audio file in the response body, which is then automatically attached to the story or pushed to your distribution channel.
This workflow works for both web and broadcast. You can generate audio for podcast episodes, radio read-alouds, or accessibility updates for the visually impaired, all triggered by the same publishing event.
- Automate audio attachment on publish
- Push to web, podcast, or broadcast
- Trigger via CMS webhook or API call
Cost Structure Overview
Pay only for what you use. The API charges $10 per 1 million characters. Spaces and punctuation count as characters, so a 1,000-character news script costs $0.01. There are no monthly fees or minimum commitments.
Failed requests are free, so you only pay for successful audio generation. Start with a free trial credit of $0.50 (about 50,000 characters) to test the quality and integration before scaling your budget.
Quality and Realism
The voices are based on licensed recordings from LibriTTS-R and LibriVox, providing natural speech patterns and clear enunciation. This provides natural speech patterns, making the audio suitable for professional broadcast.
Every generated file includes an inaudible watermark to identify it as synthetic. This ensures transparency while maintaining high production value. The voices are designed for clarity, with adjustable speed from 0.25x to 4.0x to fit your broadcast pacing.
Language Support Details
Supports 23 languages, including English, Spanish, French, German, Chinese, Japanese, and Arabic. Any voice can speak any language, allowing you to produce multilingual news broadcasts without managing multiple voice models.
The API automatically detects the language from the script if not specified. For non-English Latin-script text, you can explicitly set the language parameter to ensure accurate pronunciation. This eliminates the need for manual language selection in your workflow.
Data Privacy Aspects
Your input text is used solely to generate the audio and is not used for training models. The generated audio is returned in the response and not stored on our servers, ensuring your news scripts remain confidential.
We keep only request metadata (time, characters, status) for billing purposes. This minimizes data retention and aligns with standard editorial privacy practices. You retain full ownership of the content you send and receive.
Use Cases in News Industry
TTS APIs enable newsrooms to convert breaking news scripts into audio updates within seconds. This immediacy allows editorial teams to distribute real-time voice notifications across digital platforms without waiting for on-air talent or recording studios to become available.
By automating voiceover production, teams can scale coverage across multiple languages simultaneously. This capability ensures that urgent updates reach diverse audiences instantly, maintaining consistency and speed during high-volume news cycles.
Breaking news audio updates ['When a major event unfolds, scripts can be pushed directly to the API. The system generates audio files almost instantly, allowing editorial teams to distribute voice notifications across digital platforms without waiting for on-air talent or recording studios.', 'This immediacy ensures that breaking news reaches listeners via mobile apps or web widgets the moment text is finalized, reducing the lag between editorial decision and audience notification.'] ['Instant conversion of breaking news scripts to audio', 'Automated distribution to digital platforms and apps', 'Reduced lag between editorial decision and audience notification']
Multilingual news broadcasts ['News outlets can generate content in 23 languages using the same source text. This allows for rapid expansion into new markets without hiring specialized voice talent for each language. The API handles language detection or accepts explicit ISO codes for precise control.', 'This capability ensures that urgent updates reach diverse audiences instantly, maintaining consistency and speed during high-volume news cycles across global editions.'] ['Support for 23 languages with single-source text', 'Rapid expansion into new markets', 'Precise language control via API parameters']
Accessibility for visually impaired ['Providing audio versions of articles enhances accessibility for visually impaired readers. The synthetic voices offer a reliable alternative to inconsistent screen readers, delivering clear, consistent narration of news content. This meets growing demands for inclusive digital journalism.', 'Generated audio files are returned in standard formats like MP3 or WAV, ensuring compatibility with most accessibility tools and devices used by readers with visual impairments.'] ['Enhanced accessibility for visually impaired readers', 'Clear, consistent narration compared to screen readers', 'Compatibility with standard accessibility tools']
Technical Requirements
The NewsVoice API is designed for straightforward integration, requiring only basic HTTP knowledge. You can interact with the system using standard tools like cURL or any programming language capable of sending POST requests, making it accessible to editorial developers without deep infrastructure expertise.
The service uses standard web protocols, ensuring compatibility with existing newsroom workflows. Requests are sent via HTTPS to the API endpoint, and responses are returned in standard formats, allowing seamless connection to content management systems and automation pipelines.
Basic HTTP knowledge sufficient
The API requires only standard HTTP POST requests. Developers familiar with basic web protocols can integrate the service quickly without needing specialized middleware or complex SDKs. The official OpenAI SDKs work by adjusting the base URL, providing a familiar development experience for teams already accustomed to modern API structures.
JSON payload structure
Requests are sent as JSON bodies containing the text, voice selection, and desired output format. The structure mirrors standard speech endpoints, accepting parameters like model, input, voice, and response_format. This consistent shape allows for easy scripting and automation within existing editorial tools.
model: "tts-1" or "tts-1-hd"input: text string (up to 4096 characters)voice: one of 8 presetsresponse_format: mp3, wav, opus, flac, or pcmspeed: float between 0.25 and 4.0
Standard web protocols
Authentication is handled via a Bearer token in the Authorization header. The API uses standard RESTful conventions, returning audio files directly in the response body. This simplicity ensures reliable connectivity across various newsroom environments and reduces integration friction for large editorial teams.
Voice Selection Process
The NewsVoice API provides eight distinct voice presets, each engineered for specific editorial contexts. Newsrooms can select from female voices like Aria for clear narration or Nova for upbeat updates, and male options such as Felix for energetic promos or Atlas for neutral news delivery. Each voice is built from licensed recordings, ensuring a professional baseline.
Selection is straightforward via the voice parameter in your API request. You simply specify the preset name—like 'atlas' or 'sage'—alongside your text and model choice. The system handles the rest, delivering high-quality synthetic speech that matches your chosen tone without manual editing.
Choose from 8 presets
The API offers eight curated voice options: Aria, Nova, Luna, and Sage (female); Felix, Milo, Atlas, and Orion (male). These are not generic clones but distinct personas based on licensed LibriTTS-R and LibriVox recordings. For instance, 'Atlas' provides a neutral, news-ready tone, while 'Orion' offers a deep, documentary style.
You select the voice by passing its name in the voice field of your POST request. This allows different segments of a broadcast to have distinct auditory identities, enhancing listener engagement without requiring new voice talent hires.
- 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
Match tone to news style
Different news formats benefit from tailored vocal characteristics. Breaking news might use Felix for energy, while investigative pieces suit Orion’s depth. The speed parameter (0.25–4.0) allows further adjustment; faster speeds suit rapid updates, while slower rates aid clarity for complex data.
Note that while voices are distinct, they all speak all 23 supported languages. If using non-English text, explicitly set the language field to ensure correct pronunciation, as the system guesses from script by default.
Consistent brand voice
By standardizing on specific presets, editorial teams maintain auditory consistency across all digital and broadcast outputs. Unlike human voice actors who may vary in delivery, the API ensures every generated file from 'Atlas' sounds identical, reinforcing brand recognition.
Each audio file includes an inaudible watermark marking it as synthetic, which helps in content management without altering the listening experience. This consistency is crucial for large teams producing high-volume daily content.
image":
Integration Workflow
Integrating NewsVoice API into your editorial workflow begins with connecting your CMS via a simple webhook or middleware. Since the API follows standard HTTP protocols, you can send JSON payloads directly from your publishing environment without needing a dedicated mobile app or complex plugins. This approach ensures that your existing infrastructure remains untouched while gaining programmatic audio capabilities.
Once connected, the system listens for specific events, such as a new article being marked as 'published.' This trigger initiates the synthesis process, converting your script into audio automatically. The API handles the request, processes the text, and returns the audio file, which can then be attached to your digital asset for immediate use.
Connect to content management system
Establish a connection between your CMS and the API endpoint. Using basic HTTP knowledge, you can configure a webhook or middleware to send text data in JSON format. This setup allows your editorial team to maintain their usual publishing rhythm while the API handles the audio conversion in the background.
- Standard HTTP POST requests
- JSON payload structure
- No additional software required
Trigger audio generation on publish
Configure your CMS to trigger audio synthesis immediately upon publication. When a journalist publishes a script, the system sends the text to the API. The service processes the request, applying the selected voice and speed settings, and returns the audio file for embedding in the news update.
- Automatic trigger on publish
- Real-time processing
- Instant audio delivery
Automate distribution pipelines
Integrate the generated audio into your distribution channels. Whether it’s a podcast feed, a mobile notification, or a website broadcast, the audio file can be automatically attached to the content. This automation ensures that news updates are available in audio format across all platforms without manual intervention.
- Automatic file attachment
- Multi-channel distribution
- Scalable workflow
Cost Structure Overview
NewsVoice API charges $10 per million characters, equivalent to $0.01 per 1,000 characters. Spaces and punctuation count toward this total. Failed synthesis requests incur no charge, with the fee returned automatically.
There are no monthly subscriptions, seat licenses, or minimum spending requirements. You pay only for the audio generated, allowing precise budget control without unused capacity costs.
Pay per character used
The rate is $10 per 1 million characters. This includes spaces and punctuation. Failed requests are free.
Billing is transparent. You see exact character counts and costs in your dashboard.
- $0.01 per 1,000 characters
- Failed requests are free
No upfront costs
No monthly plans or seat licenses. No minimum spend.
Start with a $0.50 trial credit for 7 days when signing in with Google.
Scalable budget planning
Top up with USDT (TRC-20) on the TRON network. Minimum top-up is $5.
Balance does not expire. Add funds as needed for high-volume broadcast days.
Quality and Realism
NewsVoice API delivers synthetic speech grounded in licensed recordings from LibriTTS-R and public-domain LibriVox audiobooks. This foundation ensures consistent, high-fidelity audio suitable for editorial use without the variability of purely algorithmic generation.
Each generated file carries an inaudible watermark, clearly marking the content as synthetic. This provides transparency for broadcast standards and digital distribution, ensuring audiences know the voice is AI-generated while maintaining professional audio quality.
Licensed recordings as base
The voices are derived from licensed recordings, providing a natural baseline for synthetic speech. This approach avoids the robotic tone common in older TTS systems, offering clearer narration for news scripts.
Voices like Aria and Atlas provide distinct tones, from warm narration to neutral news delivery, ensuring the audio matches the editorial style of the publication.
- Based on LibriTTS-R (CC BY 4.0) recordings
- Derives from LibriVox public-domain audiobooks
- Ensures consistent audio baseline
Inaudible watermark for synthetic content
Every audio file includes a unique, inaudible watermark. This allows for easy verification of synthetic origin, meeting transparency requirements for newsrooms and digital platforms.
The watermark does not affect audio quality or listener experience, remaining transparent throughout playback.
Natural speech patterns
The API generates speech with clear enunciation and pacing. Text up to 4096 characters is processed to maintain coherence, with speeds adjustable from 0.25x to 4.0x.
Processing is sequential, ensuring consistent quality. Typical generation speed is roughly real-time for long texts, with short phrases taking a few seconds.
Language Support Details
NewsVoice API supports 23 languages, enabling global distribution without additional model training. This range covers major markets including English, Spanish, Chinese, and Arabic, ensuring broad reach for international news desks.
The system automatically detects language from the input script, defaulting to English for Latin text. For non-English Latin scripts, explicitly setting the language code ensures accuracy and prevents mispronunciation errors in broadcast-ready audio.
23 languages supported
Our API handles 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, providing flexibility for diverse editorial teams.
None
- 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 API guesses the language from the script. Latin script text defaults to English, which suits most domestic news but may require manual override for other languages using Latin characters.
Explicit language setting option
For non-English Latin-script text, set the language parameter to the ISO code. This guarantees correct pronunciation. Unsupported languages like Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, and Bengali are not currently available.
None
Data Privacy Aspects
NewsVoice API treats your editorial content with strict operational boundaries. The input text you submit for synthesis is used exclusively to generate the audio file and is not ingested into our training data. This ensures that your unpublished breaking news or exclusive scripts remain distinct from the model’s underlying voice models, preserving editorial exclusivity.
The generated audio files are returned directly in the API response and are not retained on our servers after delivery. Only the request metadata—such as timestamp, character count, and status—is stored for billing accuracy. This minimal retention policy simplifies compliance for newsrooms managing sensitive sources or embargoed stories.
Input text not used for training
Your source text is processed solely for immediate audio generation. It is not used to improve or train the underlying voice models. This separation ensures that your unique editorial voice and unpublished stories do not inadvertently influence future synthetic outputs.
This approach is particularly valuable for newsrooms handling exclusive scoops. By keeping input data separate from training datasets, you maintain full control over how your content influences the system, ensuring no unexpected voice drift or style contamination from your specific editorial corpus.
- Input used exclusively for audio generation
- No ingestion into training datasets
- Preserves editorial exclusivity
Generated audio not stored
Once the audio file is delivered to your client or system, we do not store the generated audio on our servers. This ephemeral storage model reduces data liability and ensures that your final audio assets remain under your direct control.
This immediate deletion of audio files aligns with strict privacy requirements for newsrooms that may broadcast sensitive interviews or internal briefings. You retain full ownership of the output without relying on our long-term storage infrastructure.
- Audio returned in response and deleted
- No long-term storage of generated files
- Reduced data liability
Metadata kept for billing only
We retain only the minimal metadata necessary for accurate billing and usage tracking. This includes the request timestamp, character count, and status code. No audio content or source text is kept alongside this metadata.
This limited retention policy ensures that your usage data is transparent and strictly tied to your account’s financial records. You can audit your character consumption without worrying that raw content or audio files are being retained for other purposes.
- Timestamp, character count, status kept
- No audio or text retention
- Transparent usage auditing
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
- TRON developer hub: TRC-20
- Creative Commons Attribution 4.0
- W3C WAI: making audio and video accessible
- Node.js documentation
- IETF RFC 7845: Ogg Opus
- curl documentation
- LibriTTS-R corpus (OpenSLR 141)
- LibriTTS-R paper (arXiv 2305.18802)
- LibriVox public-domain audiobooks
- OpenAI API reference: create speech
- openai-python on GitHub