Frequently asked questions
Is TTS suitable for professional training?
Yes. Voices like Sage offer calm, instructional tones ideal for corporate content. The API handles up to 4096 characters per request, allowing you to generate clear audio files for modules. It is designed for volume, letting your team focus on curriculum design while we handle the audio production efficiently.
How natural do the voices sound?
Voices are based on licensed recordings from LibriTTS-R and LibriVox. They include warm narrators like Aria or friendly conversational tones like Milo. Every file carries an inaudible watermark marking it as synthetic speech. They sound natural for training but are clearly AI-generated, not human recordings.
Can I use TTS for e-learning?
Absolutely. The API accepts text up to 4096 characters, which fits standard lesson segments. You can adjust speed between 0.25 and 4.0 to match learning paces. With 23 supported languages, you can create localized e-learning content quickly without hiring multiple voice actors for different regions.
Is it cheaper than human voiceovers?
Pricing is $10 per million characters. Failed requests are free. There are no monthly plans or minimum spends. While human voiceovers have fixed session costs, TTS scales linearly with your content volume, making it highly cost-effective for large-scale corporate training libraries and frequent updates.
How fast is the generation?
Generation is roughly real-time, about 25 characters per second. Short phrases take a few seconds. Longer texts are processed sequentially, so requests may queue under load. For most training scripts, you will receive the audio file quickly, though high volume may introduce slight delays.
Can I generate audio in multiple languages?
Yes, 23 languages are supported, including English, Spanish, German, and Japanese. Any voice can speak any supported language. If the language is not specified, it guesses from the script, but you should set the ISO code for non-English Latin text to ensure accuracy and clarity.
Do I need a developer?
No. The API endpoint matches the OpenAI speech endpoint shape. You can use the official OpenAI SDKs by changing the base URL. Alternatively, use the ready curl example provided on your dashboard. Non-technical users can integrate the audio generation into their workflows easily.
Can I update the audio easily?
Yes. Simply send a new text input via the API to generate an updated audio file. Since audio is not stored long-term, you regenerate on demand. This allows you to correct errors or update training content instantly without managing a library of static audio files.
Is the audio quality high?
The API offers two models: tts-1 and tts-1-hd. Both use the same engine and price. You can choose formats like mp3, wav, opus, flac, or pcm. The voices are clear and professional, suitable for corporate environments, though they remain synthetic with an inaudible watermark.
Can I use it for onboarding?
Yes. Onboarding materials often require consistent, clear narration. Voices like Felix (energetic) or Milo (friendly) work well for welcoming content. You can generate audio for multiple languages quickly, ensuring new hires in different regions receive clear, consistent training materials without coordinating complex voiceover schedules.
Does it support SSML?
No. The API does not support SSML, phoneme dictionaries, or custom pronunciation lexicons. You control pacing using the speed parameter (0.25–4.0) and by splitting text into smaller chunks. For precise control, you may need to adjust the input text structure manually before sending it.
Can I clone a specific voice?
No. We do not offer custom voice cloning or uploading your own voice. The eight presets are based on licensed recordings from LibriVox. You cannot create a custom clone of a specific person or brand voice for your training.
Is there a character limit?
Yes, each request accepts up to 4096 characters. Spaces and punctuation count toward this limit. For longer training modules, split your text into parts, generate the audio for each, and join the files. This ensures efficient processing and avoids exceeding the per-request constraint.
How do I choose the right voice?
Select based on tone: Aria for warm narration, Milo for conversational training, or Felix for energetic promos. Use the /v1/voices endpoint to see all options. Test different voices with sample text to find the best fit for your corporate training style and audience expectations.
Quality of Synthetic Voices
Synthetic voices for corporate training have evolved beyond the robotic tones of the past. Our models, derived from licensed LibriTTS-R recordings, produce audio that is warm, clear, and suitable for professional educational content. The speech is designed to be intelligible and engaging, ensuring learners remain focused on the material rather than distracted by unnatural cadence.
While they lack human emotion tags, the voices offer a consistent baseline of professionalism. This consistency is crucial for training modules where clarity and tone uniformity matter more than dramatic expression. The result is audio that sounds like a clear instructor rather than a machine reading a script.
Natural Sound
Our voices are based on licensed recordings from LibriVox, ensuring a natural, human-like quality without the uncanny valley effect. The audio is generated with a clear, articulate style that mimics professional narration standards found in high-quality e-learning courses.
Each voice has a distinct character, from the warm narrator to the calm instructor. This variety allows you to match the audio tone to the specific subject matter, whether it’s a soft storytelling approach for soft skills or a neutral tone for technical manuals.
- Based on licensed LibriTTS-R recordings
- Clear, articulate narration style
- Distinct character profiles for different topics
Voice Variety
We offer eight distinct voice presets, including female options like Aria and Luna, and male options like Felix and Milo. This range allows you to create diverse audio experiences without hiring multiple voice actors. You can assign different voices to different modules to keep learners engaged.
The voices support 23 languages, allowing for consistent audio quality across global training programs. Whether you need a bright, upbeat voice for onboarding or a calm, instructional tone for compliance training, there is a preset that fits the curriculum’s needs.
- 8 distinct voice presets
- Support for 23 languages
- Options for upbeat, calm, and neutral tones
Clarity
Clarity is paramount for training content. Our TTS engine is optimized for clear enunciation, ensuring that complex instructions or technical terms are heard correctly. The audio is clean, with minimal background noise or artifacts, making it suitable for both headphones and speakers.
The voices are designed to maintain a steady pace, adjustable between 0.25x and 4.0x speed. This allows trainers to slow down complex explanations or speed up simple recaps, ensuring the content is digestible for all learners, regardless of their processing speed.
- Optimized for clear enunciation
- Minimal audio artifacts
- Adjustable speed for accessibility
Cost Efficiency
SpeakWise eliminates the high fixed costs of traditional voice production. By charging $0.01 per 1,000 characters, you pay only for what you use, avoiding large upfront studio deposits.
Vs Human Voice Actors
Human voice actors charge per session or per finished minute, often requiring minimum buyouts. SpeakWise scales with your character count, making it significantly cheaper for long training modules or bulk content generation.
Vs Studio Time
Booking a studio involves scheduling, engineer fees, and post-production time. With SpeakWise, audio is generated instantly via API, removing the logistical overhead and allowing your team to focus on curriculum design rather than production logistics.
Scalability
No monthly plans or minimum spends mean you can scale from one video to thousands without renegotiating contracts. Failed requests are free, so you only pay for successful, usable audio files.
- Pay-per-character model eliminates waste
Speed of Production
Synthetic voices eliminate the wait time associated with booking studios or voice actors. You send text and receive an audio file almost immediately, turning scripts into listenable content in seconds rather than days.
Instant Generation
Requests are processed sequentially, so longer texts generate at roughly real-time speed (25 characters per second). Short phrases take just a few seconds, allowing your team to iterate quickly without blocking production timelines.
- Generate short clips in seconds
- Longer texts process at ~1x real-time speed
No Scheduling
Skip the coordination hassles of finding available talent. Since the API handles the synthesis, your curriculum designers can generate audio on-demand, anytime, without waiting for a voice actor’s schedule or studio availability.
Rapid Updates
When training content changes, you can regenerate the audio instantly. There is no need to re-book sessions or wait for post-production edits. This agility ensures your corporate training materials stay current with minimal effort.
- Regenerate audio instantly when scripts change
- No re-recording fees for minor edits
Language Support
SpeakWise supports 23 languages, including English, Spanish, German, French, Japanese, Korean, Chinese, Arabic, and Hindi. Any voice can speak any of these languages, making it easy to create multi-language training modules without hiring separate voice talent for each region.
This flexibility ensures your global workforce receives consistent, high-quality audio instruction, regardless of their native tongue. You can localize content quickly while maintaining a unified brand voice across all markets.
Multi-Language Courses
Our API supports 23 languages, including English, Spanish, German, French, Japanese, Korean, Chinese, Arabic, and Hindi. Any voice can speak any of these languages, allowing you to create multi-language training modules without hiring separate voice talent for each region.
This flexibility ensures your global workforce receives consistent, high-quality audio instruction, regardless of their native tongue. You can localize content quickly while maintaining a unified brand voice across all markets.
- Supports 23 languages including English, Spanish, German, French, Japanese, Korean, Chinese, Arabic, and Hindi.
- Any voice can speak any supported language.
- Optional language code parameter ensures accurate pronunciation.
Consistency
Use the same voice preset across all language versions to maintain brand identity. Whether you choose Aria for warm narration or Atlas for neutral news-style delivery, the tone remains recognizable even when the language changes.
This consistency helps learners feel familiar with the content, reducing cognitive load and improving retention in corporate training scenarios.
Localization
Set the language parameter explicitly for non-English Latin-script text to ensure accurate pronunciation. If omitted, the system guesses the language, defaulting to English for Latin text.
This precise control prevents mispronunciations in mixed-language scripts, ensuring your training materials sound professional and clear.
Note: Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, Bengali, and other languages are not yet supported.
- Explicit language code parameter for precise control.
- Automatic language guessing with English default for Latin text.
- Not all languages are supported (e.g., Vietnamese, Thai, Ukrainian).
Use Cases
Synthetic voices handle the heavy lifting in corporate training, turning scripts into clear audio files that keep learners engaged without the cost of studio time. By automating voiceover generation, your team can focus on curriculum design while the API handles volume.
E-Learning Modules
Generate consistent narration for entire course catalogs in hours, not weeks. With 23 languages and eight distinct voice presets, you can localize content for global teams while maintaining a uniform brand tone across all modules.
- Support 23 languages including English, Spanish, German, and Japanese.
- Use voices like Sage (calm, instructional) for structured learning.
- Process multiple scripts simultaneously via the API.
Onboarding Videos
Rapidly produce onboarding assets for new hires across different regions. Since each request takes seconds, you can update scripts and regenerate audio instantly when policies change, ensuring new employees always hear the latest information.
- Update audio instantly when HR policies change.
- Scale from 10 to 10,000 videos without hiring more voice actors.
- Maintain consistent tone across all new hire materials.
Instructional Content
Ideal for step-by-step guides where clarity matters more than emotional nuance. Voices like Felix (energetic, promo) or Atlas (neutral, news) deliver instructions clearly, making complex procedures easy to follow without distraction.
- Clear narration for technical procedures.
- Quick regeneration for updated instructions.
- No need for re-recording sessions.
Limitations
Synthetic voices deliver clear, consistent narration, but they lack the nuanced emotional depth of human performers. You cannot tag specific emotions like sarcasm or excitement, so the tone remains steady and professional. This works well for instructional content but may feel flat for dramatic storytelling.
The API accepts up to 4,000 characters per request. For longer scripts, you must split the text and join the audio files manually. While simple, this requires basic coding to handle the stitching seamlessly.
No Emotion Tags
Our voices do not support emotion tags or dynamic pitch shifts. The tone is fixed per voice preset, which ensures consistency but limits dramatic variation.
This is ideal for corporate training where clarity matters more than theatrical performance. If you need subtle emotional cues, you will need to edit the audio post-generation.
Character Limits
Each API request handles a maximum of 4,096 characters. This covers most short training modules but requires splitting for longer lectures.
You can generate multiple audio files and join them using standard audio tools. The API processes requests sequentially, so plan your generation pipeline accordingly.
Short phrases may take a few seconds to generate, while longer texts process at roughly real-time speed.
- Max 4,096 characters per request
- Requests processed sequentially
- Join audio files manually or via script
No Custom Cloning
We do not offer custom voice cloning for your team members. You choose from eight preset voices that speak all 23 supported languages.
This keeps costs predictable and avoids the complexity of training new voice models. Your team can focus on curriculum design rather than voice engineering.
If you need a specific voice, you must select the closest preset from our library.
- Eight preset voices available
- No uploading of custom voice samples
- No per-client voice design
Integration
SpeakWise integrates directly into your existing content workflows via a standard REST API. By changing the base URL to https://api.t6slotth.com/v1, you can use the official OpenAI SDKs or simple HTTP requests to generate audio from your training scripts. This eliminates the need for custom middleware or proprietary connectors.
API Integration
Our API accepts JSON bodies with text up to 4,096 characters. You can automate the generation of audio files for bulk training modules by scripting requests. The response returns the audio file directly, allowing seamless insertion into your delivery pipeline without manual file handling.
- Compatible with OpenAI SDKs via base_url change
- Standard JSON request format
- Automated batch processing support
CMS Compatibility
Since SpeakWise outputs standard audio files (MP3, WAV, etc.), they are universally compatible with most Content Management Systems. You can upload generated audio directly to articles, blog posts, or static pages without worrying about proprietary media players or plugins.
LMS Support
Learning Management Systems accept standard audio formats natively. You can attach the generated files directly to courses or modules. Note that SCORM compatibility depends on your specific LMS configuration, not the audio format itself.
Audio Formats
SpeakWise offers five distinct audio formats to ensure your training materials play seamlessly across all devices and learning platforms. The choice depends on your balance between file size, quality, and compatibility requirements.
Selecting the right format prevents playback issues and optimizes bandwidth usage for your global workforce.
MP3 Compatibility
MP3 is the universal standard for web-based video and audio players. It provides a good balance of quality and file size, ensuring your training modules load quickly on any device without buffering delays.
This format is ideal for most corporate e-learning applications where broad compatibility is prioritized over archival quality.
- Universal player support
- Efficient bandwidth usage
WAV Quality
WAV files offer uncompressed, studio-grade audio quality. This format preserves every nuance of the synthesized voice, making it suitable for high-fidelity archival or professional broadcast standards.
Choose WAV when audio integrity is critical, such as in premium executive communication modules.
- Lossless audio fidelity
- Best for archival
File Size
Audio file size scales with the response_format you select. MP3 and Opus offer significant compression, reducing storage costs for large training libraries. WAV and FLAC provide larger, high-quality files.
PCM offers raw data, resulting in the largest file sizes, while Opus provides the smallest footprint for mobile-friendly delivery.
- MP3/Opus: Smaller files
- WAV/FLAC: Larger, higher quality
Voice Selection
Choosing the right voice ensures your training materials sound professional and engaging. SpeakWise offers eight distinct presets, each designed for specific corporate contexts.
From warm narrators to neutral news readers, you can select a tone that matches your brand’s identity without needing complex configuration.
Professional Tone
Select voices like Atlas for clear, neutral news-style delivery or Sage for calm, instructional content. These presets ensure clarity in complex training modules.
- Atlas: Neutral, news-style delivery
- Sage: Calm, instructional tone
Gender Options
We provide four female and four male voices, giving you flexibility to match your team’s demographics or create diverse learning experiences.
- Female: Aria, Nova, Luna, Sage
- Male: Felix, Milo, Atlas, Orion
Style Match
Match the voice to the content style. Use Aria for warm narration or Felix for energetic promotional videos within your HR suite.
- Aria: Warm, clear narrator
- Felix: Energetic, promo style
Data Privacy
For corporate training, keeping your internal scripts secure is critical. Our API processes input text solely to generate the audio file and does not retain the content for other purposes.
Text Usage
Your training materials are used exclusively for synthesis. Once the audio is returned in the response, the text is not stored or repurposed for other outputs.
No Training Data
Unlike some consumer models, we do not use your input text to train our neural networks. Your proprietary training scripts remain yours, ensuring confidentiality for sensitive HR content.
Security
We retain request metadata—such as timestamp, character count, and status—strictly for billing accuracy. No other personal or content data is kept beyond what is necessary for the transaction.
Technical Requirements
SpeakWise is an API, meaning your team generates audio files rather than playing them directly in a browser. You need a stable internet connection to send text to our endpoint and receive the audio response. The actual playback of the resulting file (MP3, WAV, etc.) depends on your internal systems or LMS, not our API.
Internet Connection
A reliable internet connection is required for every API call. Since we process requests sequentially, a stable connection ensures your scripts submit without timeout errors. High latency may increase the wait time for longer training modules, but the final audio file is downloaded, not streamed.
- Required for submitting text and receiving audio.
- Stable connection prevents request timeouts.
- No real-time streaming dependency.
Developer Skills
Basic API integration skills are sufficient. Our endpoint mimics the OpenAI speech format, so you can use standard HTTP libraries or the OpenAI SDK with a simple base_url change. No deep audio engineering knowledge is needed to integrate our voice generation into your training workflow.
- Standard HTTP or OpenAI SDK compatibility.
- Simple JSON payload structure.
- No audio engineering expertise required.
Browser Support
Browser support depends on the audio formats you choose. MP3, WAV, and PCM are universally supported in modern browsers. If you select Opus or FLAC, ensure your playback environment handles these codecs. We do not provide a browser-based player, so your CMS or LMS handles playback compatibility.
- MP3/WAV: Universal support.
- Opus/FLAC: Codec-dependent.
- Playback handled by your CMS.
Future Proofing
SpeakWise is built as a stable API, ensuring your training infrastructure remains compatible as we make backend improvements. You do not need to worry about sudden breaking changes disrupting your automated video pipelines.
API Updates
Our API follows a consistent structure, so your integration with existing LMS or CMS platforms remains reliable. We prioritize stability for agency workflows, meaning you can schedule batch generations without fearing unexpected endpoint shifts.
New Voices
While we currently offer eight distinct presets, we may add new voice styles in the future. You can check the /v1/voices endpoint for the current list of available voices.
Language Additions
Our support covers 23 languages today. The list of supported languages may change; check the /v1/voices endpoint for the current status.
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.