How to Choose the Right Voice for Your Ad Campaign

By Rn98 editorial team · Updated 2026-11-19

This guide helps advertising agencies and radio production teams select a voice for their campaigns using the SpotVoce TTS API. We break down our eight licensed voices—from Aria’s warm narration to Orion’s deep documentary tone—so you can match your brand’s identity with precision. You will learn how to align voice characteristics with specific product types, ensuring your audio spots resonate with the intended audience.

Designed for engineers and producers, this page covers technical constraints like the 4096-character limit and speed adjustments, alongside practical workflow tips for batch generation. We also address language consistency across our 23 supported languages and provide industry-specific examples for retail, tech, and health sectors. Use this reference to streamline your creative process, reduce manual hand-offs, and deploy consistent, high-quality voice-overs programmatically.

How to Choose the Right Voice for Your Ad Campaign

Frequently asked questions

Which voice is best for TikTok ads?

Nova is a suitable choice for TikTok ads due to its bright, upbeat character. You can adjust the speed between 0.25 and 4.0 to match fast-paced editing styles. The API supports MP3 and OPUS formats, which are standard for social media platforms, ensuring quick uploads and efficient data usage for your campaigns.

What is the difference between Aria and Luna?

Aria is described as a warm, clear narrator, ideal for straightforward informational content. Luna offers a soft, storytelling tone, better suited for narrative-driven spots. Both are female voices available in the same language set. Choose Aria for clarity and Luna for emotional resonance in your audio advertisements.

Can I mix male and female voices?

The API generates one audio file per request, using a single voice preset. To mix voices, you must send separate requests for each voice and concatenate the resulting audio files. You can automate this by splitting your script into parts and joining the outputs programmatically using your preferred audio tool.

Is Felix suitable for a documentary?

Felix is energetic and promo-oriented, making it less suitable for traditional documentaries. Orion is the preferred male voice for documentaries due to its deep, authoritative tone. Felix works well for high-energy promotional spots or fast-paced commercial breaks where a dynamic, upbeat delivery is required for the audience.

How does speed affect voice quality?

Speed ranges from 0.25 to 4.0. Adjusting speed changes the playback rate without altering the pitch. Extreme speeds may affect perceived naturalness, but the engine maintains consistent quality. For standard narration, keep speed near 1.0. Faster speeds are useful for time-constrained ads, while slower speeds aid clarity in complex texts.

Which voice sounds most natural?

Naturalness is subjective. All voices are based on licensed recordings from LibriTTS-R. Each generated file includes an inaudible watermark marking it as synthetic speech. Test different voices with your specific content to determine the best fit for your brand.

Can I use Orion for news clips?

Yes, Orion is described as neutral and suitable for news-style content. His deep, documentary-style voice provides a professional tone suitable for news clips. Use the tts-1 or tts-1-hd model with Orion and set the language parameter to ensure accurate pronunciation for the specific news content you are generating via the API.

Does Sage work for educational content?

Sage is described as calm and instructional, making it ideal for educational content. Its steady tone helps maintain listener focus during longer explanations. You can adjust the speed to ensure clarity. Sage supports all 23 languages, allowing you to produce educational materials in multiple languages using the same reliable voice preset.

Is Nova too energetic for luxury brands?

Nova is bright and upbeat, which may be too energetic for some luxury brands seeking sophistication. Aria or Luna might be better alternatives for a warmer, more refined tone. However, you can lower the speed to moderate Nova’s energy. Test Nova against other voices to see if it fits your brand’s specific aesthetic.

Can I adjust the pitch of the voices?

The API does not support pitch adjustment. You can only adjust the speed (0.25–4.0) and select from eight fixed voice presets. If you need pitch variations, you must post-process the generated audio files using external audio editing software. The API returns standard formats like MP3, WAV, or FLAC for editing.

Do all voices support all languages?

Yes, every voice can speak any of the 23 supported languages. You can specify the language using an ISO code or let the system guess it from the text script. This flexibility allows you to use the same voice preset across different languages, maintaining brand consistency in your international advertising campaigns.

Which voice is best for radio?

Atlas is neutral and news-oriented, while Felix is energetic and promo-focused. For radio spots, Felix works well for commercials, and Atlas suits news updates. Luna and Aria offer warmer tones for narrative ads. Choose based on whether your radio spot requires high energy or a calm, storytelling approach for listeners.

How do I choose between tts-1 and tts-1-hd?

Both models use the same engine and price ($10 per 1 million characters). The choice depends on your quality preference. tts-1-hd and tts-1 use the same engine and price. Since the cost is identical, select tts-1-hd if you prioritize audio clarity for high-end production, or tts-1 for standard quality.

Are the voices copyrighted?

The voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and LibriVox public-domain audiobook readings. While the source material has specific licenses, the generated synthetic speech carries an inaudible watermark. You should verify the CC BY 4.0 requirements for your specific use case regarding attribution and commercial distribution.

Female Voice Options

Sound wave illustration for: Female Voice Options

SpotVoce offers three distinct female voice profiles designed to match specific campaign tones. Each voice is engineered for clarity and emotional resonance within standard advertising constraints, allowing production teams to select the optimal auditory identity without manual hand-offs.

Aria: Warm Narrator

Aria provides a warm, clear narration style ideal for brand storytelling. Her tone remains consistent across long scripts, making her suitable for voice-overs where narrative continuity is critical. She handles up to 4096 characters per request without degradation.

  • Best for: Brand narratives and storytelling.
  • Strengths: Clear enunciation and steady pacing.

Nova: Bright Upbeat

Nova delivers a bright, energetic performance suited for high-impact promotional spots. Her upbeat cadence works well for retail and e-commerce campaigns requiring immediate attention. She maintains clarity even at higher speeds, up to 4.0x.

  • Best for: Promotional spots and sales announcements.
  • Strengths: High energy and positive resonance.

Luna: Soft Storyteller

Luna offers a soft, soothing voice for storytelling segments. Her gentle tone reduces listener fatigue during longer audio spots. She is effective for lifestyle brands where a calming presence enhances message retention.

  • Best for: Lifestyle and wellness advertising.
  • Strengths: Gentle tone and narrative flow.

Male Voice Options

Sound wave illustration for: Male Voice Options

The male voice category offers distinct tonal ranges for advertising and radio production. Felix delivers high-energy delivery for promotional spots, while Milo provides a warm, approachable voice for conversational content. Atlas ensures neutral clarity for news-style reads.

Felix: Energetic Promo

Use Felix for high-impact commercials where energy drives engagement. His tone suits short, punchy scripts for retail or tech launches. The voice maintains clarity even at higher speeds, ensuring the message remains intelligible during fast-paced delivery.

  • Best for: Promo spots, limited-time offers, high-energy retail ads
  • Speed range: 1.0–2.0x for dynamic pacing

Milo: Friendly Conversational

Milo provides a warm, approachable voice for conversational content. He sounds like a trusted advisor rather than a broadcaster. This makes him ideal for explainer videos, customer onboarding, or casual brand storytelling where connection matters more than authority.

  • Best for: Explainers, onboarding, casual brand storytelling
  • Speed range: 0.8–1.2x for natural rhythm

Atlas: Neutral News

Atlas delivers a neutral, clear tone suitable for news-style reads or factual updates. His voice lacks strong emotional coloring, making it versatile for corporate communications, press releases, or informational segments where neutrality is preferred. Use him when clarity outweighs character.

  • Best for: News reads, corporate updates, factual segments
  • Speed range: 0.9–1.1x for steady delivery

Specialty Voices

Sound wave illustration for: Specialty Voices

Sage and Orion serve specific production needs.

Sage: Calm Instructional

Sage delivers a calm, instructional tone ideal for how-to videos, e-learning modules, or clear product explanations. His voice cuts through background noise without demanding attention, making it perfect for instructional overlays where clarity matters more than drama.

  • Best for: Tutorials, step-by-step guides, and technical explanations.
  • Vibe: Professional, steady, and unobtrusive.
  • Use case: Reducing cognitive load in complex product demos.

Orion: Deep Documentary

Orion provides a deep, documentary-style narration that adds gravitas to brand stories or high-end product reveals. His lower register conveys authority and sophistication, suitable for luxury spots or serious brand messaging.

  • Best for: Brand stories, luxury ads, and serious narratives.
  • Vibe: Authoritative, rich, and cinematic.
  • Use case: Elevating perceived value of premium products.

Matching Tone to Brand

Sound wave illustration for: Matching Tone to Brand

Selecting a voice that aligns with your brand’s core message ensures immediate audience recognition. In advertising, tone drives perception; a mismatch can undermine even the most compelling script. SpotVoce’s eight presets allow precise calibration of emotional resonance without manual editing.

Each voice carries distinct acoustic properties that influence listener trust and engagement. By matching vocal characteristics to your product category, you reduce cognitive load and increase message retention across video and audio spots.

High Energy Products

Felix delivers an energetic, promo-ready delivery ideal for limited-time offers, flash sales, and dynamic visual sequences. His pacing matches fast-cut editing styles common in retail and e-commerce ads.

Nova offers a bright, upbeat alternative for lifestyle brands targeting younger demographics. Both voices support speed adjustments up to 4.0x, allowing rapid delivery without losing clarity.

  • Use Felix for high-urgency calls to action.
  • Pair Nova with upbeat BGM for lifestyle spots.
  • Cap speed at 2.0x for clarity in noisy environments.

Trust and Reliability

Atlas provides a neutral, news-style delivery suitable for financial services, insurance, and B2B technology. His consistent tone builds authority without sounding salesy. Milo adds a friendly, conversational layer for customer-facing tech support or SaaS onboarding videos.

These voices minimize emotional variance, ensuring the message remains the focal point. They work well with clear, concise scripts under 4096 characters.

  • Atlas ensures neutral, authoritative delivery.
  • Milo adds approachability for consumer tech.
  • Ideal for explanatory or instructional content.

Luxury and Calm

Luna’s soft, storytelling quality suits high-end fashion, beauty, and wellness brands. Her pacing allows listeners to absorb visual details without rush. Sage provides a calm, instructional tone for premium educational content or guided experiences.

Orion’s deep, documentary style adds gravitas to luxury automotive or real estate spots. His voice conveys depth and sophistication without excessive embellishment.

  • Luna enhances visual storytelling in beauty ads.
  • Sage suits premium educational content.
  • Orion adds gravitas to luxury real estate.

Language Considerations

Sound wave illustration for: Language Considerations

Language selection directly impacts the intelligibility and emotional resonance of your ad spots. The API supports 23 languages, allowing any voice to speak any supported language. This flexibility simplifies casting for multinational campaigns without needing multiple voice talent pools.

Precise language tagging ensures accurate pronunciation and tone, especially for non-English texts using Latin scripts. Without explicit language codes, the system defaults to English, which can cause errors in other Latin-based languages like Dutch or Danish.

English Voice Nuances

Even within English, voice selection defines the brand personality. Aria provides a warm, clear narration ideal for direct-response ads, while Felix offers energetic delivery for promotional spots. Nova’s bright tone suits upbeat social media content, whereas Atlas remains neutral for news-style updates.

Ensure scripts match the voice’s natural pacing. For example, Felix handles rapid, high-energy promos well, while Luna’s storytelling style benefits from slightly slower, descriptive scripts. Speed adjustments (0.25–4.0) can fine-tune delivery but may alter emotional impact if pushed too far.

  • Aria: Warm, clear narrator for direct response.
  • Nova: Bright, upbeat for social media.
  • Felix: Energetic for promos.
  • Luna: Soft, storytelling for narratives.

Multi-Language Casting

Any of the 8 voices can speak any of the 23 supported languages. This allows consistent brand voices across markets, such as using Aria for English, Aria for French, and Aria for German campaigns. This consistency aids brand recognition across regions.

Set the language parameter explicitly for non-English Latin scripts. Without it, the system guesses English, leading to mispronunciations in languages like Dutch or Portuguese. Supported languages include Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish.

  • Supported: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish.
  • Not supported: Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, Bengali.

Accent Consistency

While voices are not tied to specific accents, their inherent tone influences perceived accent. For instance, Felix’s energetic delivery might sound more American, while Orion’s deep tone may lean British in documentary contexts. These are subjective perceptions. Test samples to ensure alignment with target demographics.

Maintain consistency across campaign assets by reusing the same voice and language settings. If switching languages, ensure the same voice is used to maintain brand identity. For example, use Atlas for news-style updates in English and German, keeping the neutral tone consistent.

  • Test samples for accent alignment.
  • Reuse voice/language settings for consistency.
  • Atlas: Neutral for news-style updates.
  • Orion: Deep, documentary tone.

Technical Constraints

Sound wave illustration for: Technical Constraints

Understanding the technical boundaries of the TTS API ensures your audio files meet production standards without unexpected failures or quality degradation.

Character Limit Impact

Each API request accepts a maximum of 4,096 characters. Longer scripts must be split into segments and joined programmatically. This limit applies to the entire input string, including spaces and punctuation.

  • Max 4,096 characters per request
  • Failed requests incur no charge
  • Split long texts before submission

Speed Adjustment Effects

The speed parameter accepts values from 0.25 to 4.0. Adjusting speed alters playback duration but may affect naturalness at extremes. Default speed is 1.0. Use speed to fit audio to specific video cuts without re-recording.

  • Range: 0.25 to 4.0
  • Default is 1.0
  • Adjust for timing fits

Format Selection

Choose mp3, wav, opus, flac, or pcm via the response_format field. MP3 is efficient for web delivery. WAV and FLAC preserve lossless quality for broadcast masters. PCM is raw data for custom processing pipelines.

  • MP3 for web efficiency
  • WAV/FLAC for broadcast quality
  • PCM for raw data access

Production Workflow

Sound wave illustration for: Production Workflow

Efficient voice-over production for ad campaigns relies on a standardized API workflow. By integrating SpotVoce into your creative pipeline, you eliminate manual hand-offs between copywriters and voice talent. This ensures that script variations are generated programmatically, maintaining consistency across multi-channel spots.

The workflow moves from script optimization to batch generation, allowing production teams to iterate quickly. With a rate limit of 60 requests per minute, you can generate multiple voice variations, though requests are processed sequentially.

Script Optimization

Limit input text to 4096 characters per request to ensure consistent audio quality. Split longer scripts into logical segments, ensuring each part forms a complete thought. This prevents awkward pauses or cut-offs in the generated audio.

Adjust the speed parameter (0.25–4.0) to match the pacing of the visual edit. Faster speeds suit high-energy retail spots, while slower rates enhance instructional content. Always verify that the adjusted speed maintains natural intonation.

  • Keep segments under 4096 characters
  • Use speed controls for pacing
  • Ensure complete sentences per segment

Batch Generation

Submit multiple requests in parallel to generate different voice variations simultaneously. This approach accelerates the selection process, allowing you to compare multiple options within minutes rather than hours.

Use the API’s consistent response format to automate audio file downloads. Integrate the generated MP3 or WAV files directly into your editing timeline, reducing manual file management overhead.

  • Parallel requests for speed
  • Automate file downloads
  • Integrate directly with editing timelines

Quality Control

Review generated audio for natural flow and tone accuracy. Use the free trial credit to test new voices before committing to bulk generation.

Check for proper language detection, especially for non-English scripts. Explicitly set the language field if the script uses Latin script but is not English, to avoid incorrect voice selection.

Verify audio formats match your post-production requirements. Choose WAV for lossless editing or MP3 for quick previews.

  • Test with free trial credit
  • Verify language settings
  • Choose appropriate audio formats

Cost Implications

Sound wave illustration for: Cost Implications

SpotVoce applies a uniform rate of $10 per million characters across all eight voice presets, ensuring predictable unit economics regardless of creative direction. This consistency simplifies budget forecasting for agencies managing multiple campaigns simultaneously.

Cost scales linearly with text volume. Spaces and punctuation count as characters, so precise script editing directly influences final invoices. Failed synthesis requests incur no charge, protecting your budget against technical errors.

Voice Price Consistency

All voices, from Aria to Orion, share the identical pricing tier. You do not pay a premium for 'HD' quality; the engine processes every request at the same base rate.

This uniformity allows you to swap voices during post-production without renegotiating unit costs, keeping your agency’s margin stable across diverse brand clients.

  • - $0.01 per 1,000 characters for all voices
  • - No tiered pricing for voice personality
  • - Failed requests are free

Length vs. Cost

Each API call accepts up to 4,096 characters. For longer scripts, split the input into chunks and join the audio files. This prevents truncation while keeping individual request sizes manageable.

Generation speed averages roughly real-time (25 characters per second). Long-form spots may queue under load, but the cost remains strictly tied to character count, not processing time.

  • - 4,096 character limit per request
  • - Spaces and punctuation count as characters
  • - Linear scaling: no volume discounts

Budget Planning

Start with the $0.50 free trial (7 days) to validate voice selection before committing to TRC-20 USDT top-ups. Minimum top-up is $5, credited automatically within minutes.

For high-volume agencies, monitor character usage per campaign. Since balance does not expire, you can accumulate credit for seasonal spikes without monthly subscription commitments.

  • - $0.50 trial credit for new accounts
  • - Minimum $5 top-up via TRC-20
  • - No monthly plans or minimum spend

Examples by Industry

Sound wave illustration for: Examples by Industry

Selecting the right voice requires matching specific tonal qualities to industry standards. For retail, energy drives conversion; for tech, clarity builds trust; for wellness, calm ensures retention. Using the correct preset reduces revision cycles and aligns audio with visual cues immediately.

Each industry has distinct pacing needs. E-commerce spots often benefit from faster speeds to convey urgency, while health content requires slower delivery for comprehension. Understanding these nuances helps agencies optimize scripts before generation.

Retail and E-Commerce

High-energy voices like Felix or Nova suit promotional spots. Use shorter scripts under 4096 characters to maintain impact. Speed adjustments between 1.0 and 1.5 can enhance urgency without sacrificing clarity.

Focus on clear calls to action. The bright, upbeat tone of Nova works well for social media ads, while Felix’s energetic style fits radio spots requiring quick brand recall.

  • Use Nova for social media ads
  • Apply Felix for radio promotions
  • Keep scripts under 4096 characters

Tech and SaaS

Clarity is paramount. Atlas provides a neutral, news-like delivery ideal for product explanations. Milo’s conversational tone works for tutorials or onboarding videos.

Ensure consistent language settings. For non-English markets, explicitly set the ISO code to avoid default guessing errors. Test different voices to find the best balance between professionalism and approachability.

  • Use Atlas for product explanations
  • Use Milo for tutorials
  • Set explicit language codes

Health and Wellness

Calm, trustworthy voices like Luna or Sage are ideal. Luna’s soft storytelling suits narrative-driven campaigns, while Sage’s instructional tone works for guided meditations or health tips.

Lower speeds (0.75–1.0) enhance comprehension. Avoid high-energy presets that might clash with the serene aesthetic. Ensure audio formats like WAV or FLAC for high-quality playback.

  • Use Luna for narrative campaigns
  • Use Sage for instructional content
  • Lower speeds for better comprehension

A/B Testing Strategies

Sound wave illustration for: A/B Testing Strategies

A/B testing isolates voice performance by running parallel ad variants. Use identical scripts and production values to ensure differences stem solely from the vocal delivery. This method provides clear data on which preset resonates with your target audience.

Track engagement metrics like click-through rates and conversion rates across variants. By comparing results from different voice presets, you can identify optimal vocal styles for specific campaign goals without guessing.

Testing Different Voices

Generate multiple audio clips using distinct presets for the same script. For example, test Aria’s warm narration against Felix’s energetic promo style. Ensure all other variables, such as speed and format, remain constant to isolate the voice as the variable.

  • Use identical script text for all voice variants.
  • Keep speed and format settings consistent across tests.
  • Test contrasting presets like Luna (soft) vs. Milo (conversational).

Measuring Engagement

Monitor key performance indicators such as click-through rates and conversion metrics. Analyze which voice variants drive higher user interaction. Data-driven decisions replace subjective preferences, ensuring your ad spend targets the most effective vocal tone.

  • Track click-through rates and conversion metrics.
  • Analyze user interaction data per voice variant.
  • Compare performance across different demographic segments.

Iterating on Scripts

Refine scripts based on voice performance data. If a voice struggles with specific phrasing, adjust the text to better suit its natural cadence. This iterative process optimizes both the script and the voice selection for maximum impact.

  • Adjust text to match voice cadence.
  • Optimize phrasing based on performance data.
  • Continuously refine scripts for better alignment.

Common Mistakes

Sound wave illustration for: Common Mistakes

Production teams often lose efficiency by ignoring technical constraints or misaligning creative choices with the API's capabilities. A mismatched voice or ignored character limits can disrupt batch generation workflows and inflate costs unnecessarily.

Understanding the specific constraints of the TTS engine ensures smoother integration into your existing creative pipeline without unexpected delays or quality drops.

Ignoring Script Length

Each request accepts a maximum of 4096 characters. Submitting a longer script causes truncation or requires manual splitting, which disrupts audio continuity.

Splitting text manually often results in visible pauses between segments. Use the API’s character limit strictly to ensure seamless audio generation.

  • Max 4096 characters per request
  • Longer texts must be split and joined
  • Spaces and punctuation count as characters

Mismatched Tone

Assigning Felix (Energetic) to a luxury brand or Aria (Warm) to a hard-sell promo creates cognitive dissonance for listeners.

Ensure the voice preset aligns with the intended emotional tone of the spot. Felix suits high-energy promos, while Orion works for documentaries.

  • Match voice personality to brand identity
  • Avoid using warm voices for hard-sell ads
  • Test tone consistency across languages

Overloading with Effects

The API does not support SSML, phoneme dictionaries, or custom pronunciation lexicons. You cannot add emphasis, pauses, or intonation tags.

Avoid assuming custom pronunciation support. If a word is mispronounced, you must adjust the input text or accept the default synthesis.

Do not expect real-time partial audio streaming; all audio is generated and returned after processing completes.

  • No SSML support
  • No custom pronunciation lexicon
  • No streaming or WebSocket support

Final Recommendations

Sound wave illustration for: Final Recommendations

Selecting the right voice is a technical decision, not just an artistic one. Use our API to generate short, 4096-character clips quickly and cheaply to validate tone before committing to full campaign scripts. This minimizes waste and ensures your creative direction aligns with the technical output.

Consistency across channels requires strict documentation. Track which voice, speed, and format combination yielded the best results for each brand vertical. This creates a reproducible standard for your production team, reducing decision fatigue and ensuring brand alignment across all spots.

Start with Short Clips

Generate 15–30 second test spots using the tts-1 model. These are fast and cost-effective. Use them to verify that the voice matches the visual pacing and emotional tone of the ad creative before scaling to longer formats.

Keep test clips under 4096 characters to avoid splitting logic. This simplifies debugging and ensures you hear the exact audio segment that will be used in the final cut.

  • Use tts-1 for rapid iteration.
  • Limit clips to 4096 characters.
  • Test with response_format: mp3 for quick previews.

Document Your Choices

Record the specific voice, speed, and language settings for every approved spot. Store these in your project management system alongside the final audio file. This ensures that any voiceover artist or automated system can reproduce the exact sound.

Note any deviations, such as speed adjustments made to fit a 30-second radio spot. This context helps future producers understand why a specific setting was chosen.

  • Log voice, speed, and format.
  • Include speed adjustments in notes.
  • Store settings with the final audio file.

Scale with Confidence

Once you have validated your voice choices, use the API’s batch capabilities to generate full campaigns. With 60 requests per minute, you can produce hundreds of spots in a single session. This allows for rapid A/B testing across different markets or platforms.

Monitor your usage and balance to ensure you have sufficient credits for large-scale deployments. The consistent pricing of $10 per million characters makes budget forecasting straightforward.

  • Generate batches of up to 60 requests/minute.
  • Forecast costs using $0.01/1000 characters.
  • Maintain a consistent balance for ongoing campaigns.

Listen to the 8 voices

Each sample below was generated by this API from the same sentence. Her gentle tone reduces listener fatigue during longer audio spots.

Avatar of the voice aria
ariafemale
Avatar of the voice nova
novafemale
Avatar of the voice luna
lunafemale
Avatar of the voice sage
sagefemale
Avatar of the voice felix
felixmale
Avatar of the voice milo
milomale
Avatar of the voice atlas
atlasmale
Avatar of the voice orion
orionmale

23 languages

Any voice speaks any of these languages.

Sources and further reading

  1. Wikipedia
  2. Python documentation
  3. OpenAI API reference: create speech
  4. FFmpeg documentation
  5. Node.js documentation
  6. LibriVox public-domain audiobooks
  7. IETF RFC 7845: Ogg Opus
  8. openai-node on GitHub
  9. curl documentation
  10. LibriTTS-R paper (arXiv 2305.18802)
  11. ISO 639 language codes
  12. openai-python on GitHub

More on this site