Frequently asked questions
What is voice cloning software?
Voice cloning software creates synthetic speech that mimics a specific person’s voice by analyzing their audio samples. It typically requires uploading recordings to train a custom model. This allows for highly personalized narration that matches the original speaker’s tone and cadence, often used for content creation or accessibility.
Is this API voice cloning software?
No. This API uses eight preset voices derived from licensed recordings, not custom cloning. You select from fixed options like Aria or Felix. Each voice is a static preset, not a unique clone of a specific individual. The voices are based on LibriTTS-R and LibriVox public-domain audiobook readings.
Can I use this for VTuber content?
Yes. The API generates audio files via a simple POST request, suitable for VTuber workflows. Voices like Luna offer soft storytelling, while Felix provides energetic promo styles. You can control speed and format, making it easy to integrate synthetic speech into your stream overlays or pre-recorded segments.
Do you offer custom voice cloning?
No. We do not offer custom voice cloning or uploading your own voice. The service relies on eight fixed preset voices. There is no voice design or emotion tagging. You cannot create a unique clone of yourself or another person; you must choose from the available presets.
How are voices licensed?
Voices are based on licensed recordings from LibriTTS-R (CC BY 4.0) and LibriVox public-domain audiobook readings. Each generated file carries an inaudible watermark marking it as synthetic speech. The voices are presets, not custom clones, ensuring clear attribution to the original source material used for training the model.
Can I clone my own voice?
No. You cannot upload your own voice to create a custom clone. The API provides eight preset voices (e.g., Aria, Felix, Atlas). You select one of these existing options for your text-to-speech needs. There is no feature to record and replicate your specific vocal characteristics.
What languages are supported?
We support 23 languages: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish. Any voice can speak any of these languages. If language is not set, it guesses from the script, defaulting to English for Latin text.
Does it support SSML?
No. This API does not support SSML (Speech Synthesis Markup Language). You cannot use XML tags to control pronunciation, pauses, or emphasis directly in the request. Control is limited to the speed parameter (0.25–4.0) and selecting a specific voice preset for your text input.
What is the cost per character?
The cost 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. There are no monthly plans or minimum spends. You pay only for the characters you successfully generate.
Is there a free trial?
Yes. Signing in with Google provides $0.50 of credit, valid for 7 days, covering about 50,000 characters. Accounts created with email and password start without trial credit. The balance does not expire, and you can top up with USDT (TRC-20) anytime for further usage.
How do I pay with USDT?
Top up using USDT on the TRON network (TRC-20) only. The minimum deposit is $5. Credits are added automatically within minutes after the transfer. We do not accept cards, PayPal, or subscriptions. There are no invoices with VAT. Send TRC-20 tokens to the address provided in your dashboard.
Can I cancel anytime?
Yes. There are no monthly plans, seats, or minimum spends. You can issue a new API key at any time, which immediately invalidates the old one. You are not locked into a contract. Your balance does not expire, so you can pause usage without losing your funds.
Is there an SLA?
No. We do not provide an SLA, uptime guarantee, or enterprise contract. There is no on-premise install option or data-center choice. Support is by email only. We do not offer certifications like SOC 2 or HIPAA. Use the API as-is, understanding that load may affect processing times.
Do you store my input text?
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 for billing purposes. Your raw text content remains transient and is not retained for model improvement.
First request
curl
curl https://api.playabcd1234.com/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.mp3Understanding Synthetic Voice Cloning
Synthetic voice cloning in TTS uses AI models trained on existing speech recordings to generate new audio from text. Unlike traditional text-to-speech, it captures the timbre and cadence of a specific voice source, allowing for consistent character voices across different scripts without re-recording.
This technology differs from traditional voice acting because it decouples performance from physical recording sessions. Instead of hiring talent for each line, developers input text into an API and receive generated audio, enabling rapid iteration for game dialogue or stream content.
What is voice cloning in TTS?
Voice cloning in TTS relies on machine learning models that analyze a dataset of human speech to replicate its acoustic properties. The system learns to map text inputs to the specific vocal characteristics of the source material, producing speech that sounds like the original speaker but reads new content.
Our engine uses licensed recordings from LibriTTS-R and LibriVox as its base. This means the 'cloned' voices are derived from public-domain audiobooks and licensed corpora, providing a stable, legally clear foundation for synthetic speech generation without relying on a single celebrity's likeness.
- Models learn acoustic patterns from source recordings
- Generates new speech from text input
- Maintains consistent vocal timbre across scripts
How it differs from voice acting
Traditional voice acting requires a human performer to interpret scripts in real-time, capturing nuance through breath, timing, and physical effort. Synthetic speech replaces the performer with an algorithm that applies learned vocal traits to text, ensuring consistency but lacking the spontaneous emotional variance of human acting.
While voice acting offers unique interpretative choices, synthetic voices provide scale and reproducibility. For VTubers, this means a character's voice remains identical across hundreds of videos, regardless of the original voice actor's availability or fatigue levels.
Why VTubers use synthetic audio
VTubers and streamers use synthetic audio to maintain character consistency across global audiences and diverse content formats. By using preset voices like Aria or Atlas, creators can generate dialogue in multiple languages without hiring separate voice actors for each region.
This approach reduces production costs and time significantly. A streamer can input script updates instantly, allowing for dynamic interaction with chat or rapid content creation without scheduling studio time. It also enables easy A/B testing of voice tones for different video segments.
Preset Voices vs. Custom Cloning
EchoVocal delivers broadcast-ready synthetic audio through eight distinct preset voices, each designed for specific streaming contexts. These presets range from the warm narration of Aria to the energetic promo style of Felix, ensuring your VTuber content maintains a consistent, professional tone without the variability of live performance.
Available preset voices
We offer eight licensed presets: Aria, Nova, Luna, and Sage for female voices; Felix, Milo, Atlas, and Orion for male voices. Each voice is distinct, allowing you to switch personas for different segments or character roles. All voices support 23 languages, enabling global reach without needing separate voice actors.
- Aria: Warm, clear narrator
- Felix: Energetic, promo
- Atlas: Neutral, news
No custom voice cloning for customers
Our platform focuses on high-quality presets rather than custom voice cloning. You cannot upload your own audio to create a unique voice model. This ensures consistent performance and avoids the ethical complexities of cloning, keeping the workflow simple for indie developers who need reliable, immediate audio generation.
Licensing: LibriTTS-R and LibriVox
Every voice is built from recordings licensed under LibriTTS-R (CC BY 4.0) and sourced from LibriVox public-domain audiobooks. This foundation means your generated audio is legally safe to use. We also embed an inaudible watermark in every file to clearly mark it as synthetic speech, respecting creator rights and transparency.
Consent and Ethics in Synthetic Speech
Ethical synthetic speech begins with transparent sourcing. Our voices are built from LibriTTS-R, derived from LibriVox public-domain audiobook readings licensed under CC BY 4.0. This foundation ensures that the base recordings are legally cleared, providing a clear lineage for content creators who prioritize provenance in their VTuber or streaming workflows.
Source material: public domain audiobooks
We use licensed recordings from LibriVox, which are in the public domain. The underlying audio data carries a CC BY 4.0 license, allowing broad commercial and non-commercial use without complex royalty tracking. This means you can generate voice lines for your indie game or stream without fearing copyright strikes on the base voice model.
Watermarking synthetic audio
Every generated audio file includes an inaudible watermark that explicitly marks it as synthetic speech. This invisible tag allows platforms and audiences to identify AI-generated content automatically. It provides a layer of transparency, helping maintain trust with your viewers by ensuring your synthetic voice is clearly distinguished from human recordings.
Creator rights and attribution
Since the source material is public domain, you retain full rights to the generated audio clips you produce. There are no ongoing royalty obligations for the voice model itself. However, we recommend attributing the source if required by your specific platform or community guidelines, as a best practice for ethical content creation.
Technical Specs for Streamers
The API accepts mp3, wav, opus, flac, and pcm formats via the response_format field. This flexibility allows streamers to choose lossy compression for bandwidth efficiency or lossless WAV for archival quality without re-encoding.
Generation typically runs at roughly real-time speed (25 characters per second). Requests process sequentially, so queueing may occur under load, making pre-generation ideal for scheduled stream segments.
Supported audio formats (mp3, wav, opus)
Select your preferred format in the request body. MP3 and Opus are efficient for live streaming overlays, while WAV and FLAC provide uncompressed clarity for VODs. PCM is available for raw data processing.
- MP3: Standard for web streaming
- WAV: Uncompressed, high quality
- Opus: Efficient codec
- FLAC: Lossless audio storage
Latency and real-time generation
Synthesis takes approximately 25 characters per second for long texts. Short phrases may take a few seconds to process. Because requests are queued sequentially, high traffic can introduce slight delays compared to pre-recorded assets.
Speed control for live streams
Adjust playback speed from 0.25x to 4.0x using the speed parameter. This allows streamers to sync synthetic dialogue with on-screen actions or compress long narrations for fast-paced VTuber segments without altering pitch artificially.
Voice Cloning Software Comparison
While the OpenAI speech endpoint shares the same request shape, our API offers distinct pricing and format flexibility. The OpenAI model costs $0.015 per 1,000 characters for standard and $0.03 for HD, whereas our API charges a flat $0.01 per 1,000 characters for both tts-1 and tts-1-hd models. We also support Opus and FLAC outputs natively, which are essential for VTuber audio pipelines seeking high compression or lossless quality without extra conversion steps.
Unlike ElevenLabs, which often requires subscription tiers for access to certain voice features, our model is purely usage-based with no monthly minimums. We provide eight high-quality preset voices derived from licensed LibriTTS-R recordings, ensuring ethical sourcing without the need for custom voice cloning infrastructure. This approach keeps overhead low for indie developers who need reliable, broadcast-ready audio without managing complex tier structures.
OpenAI vs. our API
The endpoint structure mirrors OpenAI's, allowing you to swap the base URL to https://api.playabcd1234.com/v1 and use existing SDKs. However, our pricing is strictly $0.01 per 1,000 characters for both models, compared to OpenAI's tiered pricing of $0.015 and $0.03. We also include Opus and FLAC formats directly, which OpenAI does not support natively in their speech endpoint.
- Flat $0.01 per 1k chars for all models
- Native Opus and FLAC support
- Same JSON request shape as OpenAI
ElevenLabs alternative features
We offer eight curated voices (Aria, Nova, Luna, Sage, Felix, Milo, Atlas, Orion) licensed from LibriVox public-domain audiobooks. Unlike services requiring custom cloning subscriptions, we provide immediate access to these voices with no tier restrictions. Each generated file includes an inaudible synthetic watermark, ensuring clear attribution of AI-generated content for streamers concerned with transparency.
- 8 preset voices from licensed sources
- No subscription tiers required
- Inaudible synthetic watermark included
Cost per character analysis
Our rate is $10 per million characters, equating to $0.01 per 1,000 characters. This is lower than most competitors offering similar voice quality. Failed requests are free, ensuring you only pay for successful synthesis. With a free trial of $0.50 credit (approx. 50,000 characters), you can test the audio quality before committing to USDT top-ups.
- $0.01 per 1,000 characters
- Free failed requests
- $0.50 free trial credit

Getting Started with StreamVoice
Start by signing in via Google on the /get-api-key/ page to receive $0.50 in trial credit (approx. 50,000 characters) valid for seven days. Email-based accounts begin without this bonus. Once registered, your dashboard displays your API key, current balance, and usage statistics alongside a ready-to-use cURL example for immediate testing.
Top-ups require USDT on the TRON network (TRC-20) with a $5 minimum. Transfers are credited automatically within minutes, with no cards or subscriptions needed. Your balance never expires, and you can issue a new API key at any time, instantly invalidating the previous one for security.
Google sign-up and free trial
Sign in with Google on the /get-api-key/ page to instantly receive $0.50 of credit, covering about 50,000 characters for seven days. Accounts created with email and password do not receive this trial credit. The registration page immediately shows your active API key, current balance, and a working cURL example for testing the speech endpoint.
USDT top-up process
Add funds exclusively via USDT on the TRON network (TRC-20). The minimum top-up is $5, and credits are applied automatically within minutes of transfer. There are no monthly plans, invoices, or VAT charges. Your balance does not expire, allowing you to pay only for the synthetic audio you actually generate.
API key management
Each account holds one active API key, displayed clearly on the dashboard alongside your usage metrics. You can issue a new key at any time; the old key stops working immediately, ensuring tight control over your API access. Use this key with the standard Authorization header in your requests.
Language Support for Global Audiences
EchoVocal’s API supports 23 languages, enabling VTubers to reach global audiences without managing separate voice models. The system accepts texts in Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, and Turkish.
23 supported languages
All eight preset voices can speak any of the 23 supported languages. This flexibility allows a single voice model to serve diverse character roles across different regional broadcasts.
- Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Russian, Spanish, Swahili, Swedish, Turkish
Auto-detection vs. manual setting
If the language parameter is omitted, the API guesses the language from the script. Latin script defaults to English, but for non-English Latin text, you must explicitly set the language code to ensure accurate pronunciation.
Limitations: no Vietnamese or Thai
The API does not support Vietnamese, Indonesian, Thai, Ukrainian, Romanian, Czech, Bengali, or other languages outside the 23 listed. If your VTuber content relies on these languages, the current preset voices will not handle them correctly.
Data Privacy for Indie Developers
Indie developers often worry about whether their narrative text becomes part of a public model. With EchoVocal, input text is used solely to generate the audio file for that specific request. Once the audio is returned, it is not stored on our servers, ensuring your unique scripts remain transient and private.
Input text usage policy
The text you send to the API is processed only to create the corresponding audio output. It is not archived or reused for other customers’ generations, keeping your story assets distinct from the general pool of synthetic media.
No training on your data
Your submitted text is never used to train or improve our underlying voice models. The synthetic voices are based on fixed licensed recordings from LibriTTS-R and LibriVox, so your specific input does not influence future voice quality or character generation.
Metadata retention for billing
We retain minimal metadata—specifically the timestamp, character count, and request status—strictly for accurate billing and rate limit tracking. This data is used only to calculate your charges and manage your API key’s usage quota.
Frequently Asked Questions
Clarifying the distinction between our preset engine and true cloning helps set accurate expectations for streamers seeking consistent, ethical audio assets.
Is this voice cloning software?
No. We offer eight high-quality preset voices (e.g., Aria, Felix) rather than custom cloning. While these voices sound natural, they are based on licensed recordings, not a copy of your specific voice.
Can I clone my own voice?
Not currently. We do not upload or process your voice for cloning. You select from our eight available presets, which can speak all 23 supported languages.
Does it support SSML?
SSML is not supported. You control pacing via the speed parameter (0.25–4.0) and split text manually if needed. Emotion tags are also unavailable.
Listen to the 8 voices
Each sample below was generated by this API from the same sentence. VTubers and streamers are turning our indie games into viral sensations, driving massive engagement through their unique and entertaining content.
23 languages
Any voice speaks any of these languages.
