Frequently asked questions
Is this API better than GCP Speech API for news?
News Audio Wire offers a transparent, character-based pricing model at $10 per million characters, avoiding complex enterprise tiers. Unlike many competitors, we provide a simple REST API compatible with OpenAI SDKs, making integration straightforward for editorial teams scaling audio content without hidden fees or minimum spend requirements.
How does the cost compare to Google Cloud TTS?
Our rate is $10 per million characters, or $0.01 per 1,000 characters. This flat rate applies to all models and voices, with no monthly commitments or seats. Failed requests are free, ensuring you only pay for successful audio generation, simplifying budget forecasting for high-volume news pipelines.
Can I use this for automated news broadcasts?
Yes. The API accepts up to 4096 characters per request, allowing you to split longer scripts and join audio files. With a speed range of 0.25–4.0, you can optimize generation time. The straightforward REST endpoint integrates easily into existing automated workflows for consistent, scalable audio output.
What voices are best for news reporting?
For news reporting, Atlas offers a neutral, professional tone suitable for hard news. Milo provides a friendly, conversational style for interviews or lighter segments. Both voices support all 23 languages, ensuring consistent quality across global news desks without needing to switch voice models for different regions.
Do I need to pre-purchase credits?
No. You can sign up with Google to receive a $0.50 free trial valid for 7 days. For ongoing use, top up via USDT (TRC-20) with a minimum of $5. Balances do not expire, and you can issue a new API key at any time without service interruption.
How many languages are supported?
We support 23 languages, including English, Spanish, German, French, Japanese, and Hindi. Any voice can speak any supported language. If the language parameter is omitted, the system guesses from the script, defaulting to English for Latin text. Vietnamese, Thai, Indonesian, Ukrainian, Romanian, and Czech are not supported.
Is there a limit on characters per request?
Each API request accepts a maximum of 4096 characters for the input text. For longer news articles, split the text into parts of this size and join the resulting audio files. This ensures consistent performance and allows you to manage processing load effectively across your editorial workflow.
Do you offer SSML support for news?
No, we do not offer SSML support. The API uses simple text input with optional language codes and speed adjustments. This simplifies integration for automated systems, as you do not need to manage complex markup structures, making it easier to feed raw text from your CMS directly to the API.
Can I clone a reporter's voice?
No, we do not offer custom voice cloning. Our eight voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and LibriVox public-domain audiobook readings. You select from presets like Aria, Nova, or Atlas. Each generated file includes an inaudible watermark to identify it as synthetic speech, ensuring transparency in your audio content.
Is the service reliable for live news?
The service processes requests sequentially, meaning under load, requests may wait in a queue. Typical speed is roughly real-time for long texts, but short phrases may take a few seconds. We do not offer an SLA or uptime guarantee, so plan for potential latency during peak news hours.
What happens if a request fails?
If synthesis fails, the charge is returned in full. You are not billed for unsuccessful attempts. This ensures that network issues or processing errors do not impact your balance. You can retry the request with the same text and parameters, or adjust settings like speed or voice.
Do I get a refund for failed requests?
Yes. Failed requests are free, and any charge incurred is returned in full. This policy applies to all request types. You only pay for successfully generated audio files, making it cost-effective for testing different voices or languages before committing to large-scale news production workflows.
Can I use this with my existing CMS?
Yes. The API uses a standard JSON body compatible with the OpenAI SDKs. You can integrate it by changing the base URL to https://api.d1555.info/v1. This allows you to connect your existing CMS or automation tools without needing a custom SDK or complex middleware for audio generation.
How long does synthesis take for news?
Synthesis takes roughly real-time for long texts, generating about 25 characters per second of audio. Short phrases may take a few seconds. Requests are processed sequentially, so under heavy load, a request may wait in a queue. Plan your generation timing accordingly for time-sensitive news broadcasts.
First request
curl
curl https://api.d1555.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.mp3The Cost Problem in News Audio
Newsrooms generate audio at scale, making per-seat enterprise licenses inefficient. Our API charges strictly per character ($0.01 per 1,000 characters), eliminating unused capacity costs. This model aligns directly with your content volume, ensuring editorial budgets cover only what is actually broadcast.
Per-character billing vs. per-seat models
Traditional enterprise TTS solutions charge for seats or monthly tiers, regardless of usage. News desks often pay for peak capacity during off-peak hours. Our character-based pricing ensures you pay only for the text processed, offering predictable costs for variable news cycles.
- Pay $0.01 per 1,000 characters.
- No minimum monthly spend or seats.
- Failed requests are free.
How GCP pricing scales with high-volume news feeds
High-volume feeds can trigger unexpected cost spikes with tiered or per-request pricing models. Our flat rate of $10 per million characters provides stability. Whether you process 10,000 characters or 10 million, the unit cost remains constant, simplifying financial forecasting for large teams.
Transparent costs for editorial budgets
Every character counts, including spaces and punctuation. The dashboard shows real-time usage, balance, and characters left. There are no hidden fees, VAT invoices, or subscription traps. You can top up with USDT on the TRON network (TRC-20) starting at $5, with credits credited automatically within minutes.
Technical Compatibility for Newsrooms
Newsrooms already familiar with OpenAI’s TTS integration can adopt this API with minimal friction. By adjusting the base_url to https://api.d1555.info/v1, teams retain their existing code structure while gaining access to a character-based pricing model better suited for high-volume news feeds.
OpenAI SDK compatibility for existing pipelines
The endpoint mirrors the OpenAI speech structure, allowing immediate use of official SDKs. This eliminates the need for custom HTTP clients or complex middleware when migrating from legacy enterprise solutions.
- Drop-in replacement for OpenAI TTS endpoints
- Maintains existing integration logic
Standard JSON endpoints for easy integration
Integration relies on standard POST requests with JSON bodies, ensuring compatibility with most modern CMS and automation tools. The consistent request shape simplifies debugging and reduces implementation time for engineering teams.
Support for MP3, WAV, and Opus formats
Audio outputs are available in MP3, WAV, Opus, FLAC, and PCM, covering requirements for web players, broadcast archives, and mobile apps. This flexibility ensures optimal file sizes and quality for diverse distribution channels.
Voice Quality for News Content
News audio demands clarity over performance. Our neutral voices deliver straight reporting without distracting inflection, ensuring the journalism remains the focus. Editors can choose between Atlas for standard news and Orion for deeper, documentary-style features.
For longer formats, storytelling voices like Luna or Sage bring consistent warmth to features and podcasts. Each voice is derived from licensed recordings, offering a distinct character that scales seamlessly across your content library without requiring new talent hires.
Neutral voices for straight news reporting
Atlas provides a neutral, news-ready tone ideal for breaking updates and factual briefings. Orion offers a deeper, documentary-style delivery suitable for investigative pieces or long-form analysis. Both ensure the voice remains transparent to the content, avoiding the artificial enthusiasm often found in standard TTS.
These presets maintain consistent pacing and articulation, crucial for listener comprehension during high-volume news cycles. Whether reading a ticker or a detailed report, the audio remains clear and professional, matching the editorial standards of major publications.
- Atlas: Neutral, clear, suitable for standard news reporting
- Orion: Deep, documentary-style, ideal for investigative pieces
Storytelling voices for features and podcasts
Luna offers a soft, storytelling tone perfect for narrative features and human-interest stories. Sage provides a calm, instructional voice ideal for explainer segments or educational news content. These voices add emotional depth without sacrificing clarity, enhancing listener engagement for longer audio pieces.
Unlike generic synthesis, these presets have a distinct character that helps differentiate your content. They are based on licensed recordings, ensuring a natural flow that supports the journalistic narrative without drawing excessive attention to the voice itself.
- Luna: Soft, storytelling tone for narrative features
- Sage: Calm, instructional voice for explainer segments
Consistent audio across 23 languages
All eight voices support 23 languages, including Arabic, Chinese, French, German, Japanese, Korean, Spanish, and more. This ensures tonal consistency across your global editions, whether you are producing content for European or Asian markets. The same voice character remains recognizable regardless of the language spoken.
Automatic language detection works for Latin scripts, but explicit language codes are recommended for non-English Latin text to ensure accuracy. This flexibility allows editorial teams to scale content production across regions without managing multiple voice engines or inconsistent audio qualities.
- Support for 23 languages including major European and Asian markets
- Consistent voice character across all supported languages
Scaling for Editorial Teams
Newsrooms generate content at varying velocities, from breaking alerts to deep-dive features. Our API handles this volume by splitting input text into 4,096-character chunks, synthesizing each segment individually before joining the audio files. This chunking strategy ensures that large daily briefings do not timeout or exceed payload limits, allowing editorial teams to process extensive scripts in a single operation.
Handling 4096 character chunks efficiently
Each request processes up to 4,096 characters, covering roughly 15–20 seconds of audio. For longer articles, the system splits the text, generates the parts, and stitches them together seamlessly. This approach keeps individual request sizes manageable while delivering complete, coherent audio files for your CMS or broadcast pipeline.
- Input limit: 4,096 characters per request
- Automatic joining of segmented audio files
- Supports standard formats: mp3, wav, opus, flac, pcm
Queue management during peak news cycles
Requests are processed sequentially, meaning high-volume periods may introduce a brief queue. For most news content, generation speed is roughly real-time (25 characters per second). During breaking news surges, short phrases may take a few seconds longer, but the system prioritizes completion over sub-second latency, ensuring no audio is dropped.
No minimum spend for small or large desks
There are no monthly seats or minimum spend requirements. You pay only for what you use, with 60 requests per minute allowed per key. This flat-rate structure scales effortlessly whether your desk generates five briefs a week or five hundred, making it ideal for teams with fluctuating editorial cycles.
- 60 requests per minute per key
- No monthly plans or seats
- Transparent $10 per 1M characters billing
Language Support for Global News
News Audio Wire supports 23 languages, covering major markets from Arabic and Chinese to English and Spanish. This allows publications to produce localized content without managing separate voice models for each region.
All voices can speak any of these languages, ensuring consistent audio quality whether you are generating a local bulletin or an international dispatch. The engine handles diverse linguistic structures without requiring specialized inputs.
Coverage of major European and Asian languages
The API covers 23 languages, including English, Spanish, French, German, Italian, Portuguese, Russian, Chinese, Japanese, Korean, Arabic, Hindi, and others. This range supports global news desks producing content for diverse audiences without needing multiple specialized providers.
Key languages like Dutch, Polish, Swedish, and Turkish are fully supported, ensuring that regional editions of your publication sound natural and accurate. We do not support Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, or Bengali.
- 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 vs. explicit settings
If the language field is omitted, the system guesses the language from the text script. Latin-script text defaults to English, which is suitable for most English news but may misinterpret other Latin-based languages.
For non-English Latin-script text, such as French or German, you should explicitly set the language parameter to ensure accurate pronunciation. This prevents the engine from defaulting to English and producing awkward results.
Handling non-Latin scripts accurately
The engine correctly processes non-Latin scripts including Arabic, Chinese, Hebrew, Hindi, Japanese, Korean, and Russian. Each voice adapts to the phonetic requirements of the specified language, maintaining clarity in complex scripts.
Consistent voice quality is maintained across all 23 supported languages, allowing editorial teams to switch between languages seamlessly within the same workflow without losing audio consistency.
Data Privacy for Journalists
When your editorial team sends breaking news or exclusive interviews to the API, that input text is used strictly to produce the audio file. Once the synthesis is complete, the generated audio is returned in the response and not stored on our servers. Your actual news content does not linger in storage after delivery.
Input text used only for generation
We treat your input as a transient command. The text you submit is processed solely to generate the specific audio file you requested. There is no secondary use or archival of the raw text payload beyond the immediate creation of the audio asset.
No training on your news content
Unlike some platforms that build voice models from user data, we do not use your submitted text to train our engines or improve future voice synthesis. Your journalistic content remains yours, ensuring that exclusive scoops or unpublished drafts are never ingested into our training datasets.
Metadata retention for billing clarity
We retain request metadata—specifically the timestamp, character count, and status—for billing purposes. This allows you to verify charges against your usage exactly. Since we bill per character, this metadata provides the transparency needed for editorial budget reconciliation without storing the actual news content.

Migration from Legacy APIs
Migrating from legacy enterprise speech APIs is straightforward because our endpoint mirrors the OpenAI speech structure. You only need to update the base_url in your existing OpenAI SDK configuration to https://api.d1555.info/v1 and swap in your new API key. This minimal change allows editorial engineering teams to redirect their current pipelines without rewriting core integration logic.
Our SDK compatibility ensures that standard JSON payloads for model selection, voice assignment, and format output work immediately. This reduces the friction of moving from per-seat enterprise contracts to our transparent character-based billing, keeping your engineering time focused on content rather than infrastructure debugging.
Switching base_url in OpenAI SDK
Update your SDK configuration to point to https://api.d1555.info/v1 using your unique API key. The request shape remains identical to the OpenAI standard, so your existing code for handling responses and errors continues to work without modification.
Mapping existing voice presets to new ones
Replace your legacy voice identifiers with our eight presets. For straight news reporting, switch to atlas (neutral) or sage (calm, instructional). For features or podcasts, use aria (warm, clear narrator) or luna (soft, storytelling).
- atlas: Neutral, news-ready
- sage: Calm, instructional
- aria: Warm, clear narrator
- luna: Soft, storytelling
Testing with free trial credit
Sign in with Google to receive $0.50 in trial credit, valid for 7 days. This provides approximately 50,000 characters for testing your integration with real news text before committing to a TRC-20 top-up.
Real-World Use Cases
Newsrooms often juggle multiple audio pipelines for different distribution channels. This API integrates directly into existing editorial workflows, allowing teams to generate audio from the same text sources used for web articles and newsletters without managing separate enterprise contracts or per-seat licenses.
By using standard OpenAI SDK compatibility, engineers can swap the base URL and continue using familiar code. This reduces the technical overhead for newsrooms that already have automation scripts in place, enabling quick deployment of audio features without hiring specialized integration staff.
Automated breaking news audio updates
When a major story breaks, editorial teams can trigger a script to convert the latest wire text into audio. The API processes requests in a queue, ensuring that even during high-volume news cycles, each request is handled sequentially. This allows for real-time audio updates that mirror the text feed, keeping listeners informed without manual recording.
Since each request is limited to 4096 characters, longer breaking news updates are split and stitched together automatically. This ensures consistent audio quality and length, preventing the audio file from becoming too large for efficient streaming or downloading by subscribers.
- Trigger audio generation from CMS publish events
- Automatic splitting of long wire text into audio segments
- Sequential processing ensures stable load during news spikes
Daily briefing generation for subscribers
News publications can automate the creation of daily audio briefings. By feeding the day's top stories into the API, editorial teams can produce a consistent audio product for subscribers. The neutral voices, such as Atlas for news, provide a professional tone suitable for morning or evening digest formats.
The transparent per-character pricing means costs scale directly with the volume of subscribers and content length. There are no minimum monthly fees, making it viable for both large publications and niche newsletters to offer audio content without significant upfront investment or risk.
- Generate consistent daily audio digests
- Scale costs directly with subscriber volume
- No monthly minimums for small or large desks
Multilingual news dispatches
For global news desks, the API supports 23 languages, allowing teams to dispatch news in multiple languages from a single text source. The automatic language detection simplifies the process, though explicit language codes ensure accuracy for non-Latin scripts or ambiguous text.
This capability enables newsrooms to expand their reach without hiring voice talent for each language. The consistent voice presets across languages provide brand recognition, whether the content is a local update or an international report, maintaining a uniform editorial voice across all markets.
- Support for 23 languages including Asian and European
- Automatic language detection with explicit override option
- Consistent voice branding across global markets
Why This API Fits Newsrooms
We prioritize clear, neutral voices like Atlas for straight news and Orion for documentaries, ensuring factual delivery without emotional bias or celebrity imitation. The API returns synthetic audio with an inaudible watermark, providing transparency about content origin without visual clutter.
Editorial focus on factual reporting
Our voice presets are designed for clarity and neutrality. Atlas delivers a neutral, news-ready tone, while Orion provides a deep, documentary style suitable for in-depth features. We avoid emotion tags, ensuring the voice remains a transparent vessel for the text.
Inaudible watermark for synthetic content
Every generated audio file includes an inaudible watermark marking it as synthetic speech. This allows newsrooms to maintain transparency with their audience regarding automated content, distinguishing it from human-recorded broadcasts without requiring on-screen labels.
Simple email support for technical issues
We offer direct email support for technical integration questions. Support is by email only; our team handles API key issues, rate limit inquiries, and format questions via email.
FAQ
News teams often seek predictable costs and straightforward integration. Our character-based billing and OpenAI SDK compatibility address these needs directly, offering a transparent alternative to legacy enterprise models.
Is this API better than GCP Speech API for news?
It depends on your volume and infrastructure. We offer a simpler, character-based pricing model without per-seat fees, which often reduces costs for high-volume news feeds. Integration is streamlined via standard OpenAI SDK compatibility.
How does the cost compare to Google Cloud TTS?
We charge $10 per 1 million characters, regardless of voice or format. This flat rate often undercuts complex tiered pricing structures found in other cloud providers, especially for long-form content.
Can I use this for automated news broadcasts?
Yes. The API accepts text inputs up to 4096 characters, allowing you to split longer scripts and join audio files programmatically for seamless automated updates.
What voices are best for news reporting?
Atlas provides a neutral, news-style tone suitable for straight reporting. Millo offers a friendly, conversational style ideal for commentary or lighter segments.
Do I need to pre-purchase credits?
No. You can start with a $0.50 free trial. Top-ups via USDT (TRC-20) are minimum $5 and credit automatically. No subscriptions or minimum spend requirements.
How many languages are supported?
We support 23 languages, including English, Spanish, German, French, Japanese, and Arabic. Any voice can speak any supported language.
Is there a limit on characters per request?
Yes, each request is limited to 4096 characters. For longer texts, split the input and join the resulting audio files.
Do you offer SSML support for news?
No. We do not support SSML, phoneme dictionaries, or custom pronunciation lexicons. Input is plain text.
Can I clone a reporter's voice?
No. We do not offer custom voice cloning. Voices are based on licensed recordings from LibriVox and LibriTTS-R.
Is the service reliable for live news?
Requests are processed in queue order. During peak news cycles, generation may take slightly longer than real-time, but failed requests are free.
What happens if a request fails?
If synthesis fails, the charge is returned in full. Your balance remains unaffected.
Do I get a refund for failed requests?
Yes. Failed requests are free, and any charged amount is returned to your account immediately.
Can I use this with my existing CMS?
Yes. Since we follow the OpenAI speech endpoint shape, you can integrate via standard HTTP or the OpenAI SDK by adjusting the base URL.
How long does synthesis take for news?
Typically, generation takes about 25 characters per second. Short phrases may take a few seconds; longer texts are processed sequentially.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. Requests are processed sequentially, meaning under heavy load, a request may wait in a queue before synthesis begins.
23 languages
Any voice speaks any of these languages.
