Text to speech
Robivox
Robivox’s access terms set a registration gate for longer text and cap each synthesis request at 5,000 characters. The provider says almost all voices may be used commercially, while its offer requires a separate license agreement for commercial use of the voice models it names. That distinction matters before choosing a voice for work intended for commercial release.
These terms are most relevant to buyers weighing short-form audio production against text length, token conditions, and voice rights.
Editorial review based on public sources, without hands-on testing; facts checked on 2026-10-31.
This page contains no partner link to this service.
- Name
- Robivox
- Category
- Text to speech
- Official site
- robivox.ru · No link provided
- Facts checked
Robivox’s synthesis limits set the entry terms
The provider’s homepage, offer, and FAQ describe the service as follows, checked 2026-10-31.
- Service: AI speech synthesis that creates audio files (offer, checked 2026-10-31).
- Stated uses: videos, IVR menus, and automated calls (homepage, checked 2026-10-31).
- Downloads: MP3 or WAV files (homepage, checked 2026-10-31).
- Unregistered synthesis: up to 100 characters (FAQ, checked 2026-10-31).
- Longer text: registration required (homepage, checked 2026-10-31).
- Request ceiling: 5,000 characters (homepage, checked 2026-10-31).
Bonus rubles meet expiring tokens
The provider says registration comes with 5 bonus rubles, while access can also be obtained by paying under its current published tariffs. The FAQ describes the bonus as credit for testing the service.
The FAQ gives approximate synthesis examples rather than a full tariff schedule: it says 100 rubles buys approximately 100 minutes with a regular voice or 20 minutes with a PRO voice. It also says 100 rubles buys approximately 1000 characters of synthesis with PRO+ voices. The provider states these as examples; the PRO+ example is character-based, not time-based.
Token balances expire 180 calendar days after crediting, and unused tokens are non-refundable except where Russian law expressly provides otherwise, according to the offer. The homepage says PRO+ voices cannot be used with bonus tokens, so those voices require a balance top-up.
The FAQ lists commercial use of voices by Natalia Vershinina and Matvey Nikoloidi as starting at 1000 ₽. The offer separately requires a license agreement for commercial use of the voice models it names.
Voice models turn text into files
Robivox says users choose a voice model and synthesis parameters before creating speech from their text. Its homepage describes a voice as AI-created from a real recorded speaker’s speech.
The FAQ says users can mark stress by putting a plus sign before the stressed vowel. It also states that multilingual voice models can synthesize text in any language.
The service offers text-to-speech for videos, IVR menus, and automated calls, with synthesized files available for download in MP3 or WAV.
Registration gates longer synthesis
The homepage and FAQ set the registration threshold at text longer than 100 characters; unregistered users can synthesize text up to that length. A single request is capped at 5,000 characters, including spaces, according to the homepage and offer.
In practice, longer text requires an account, and even registered users must split anything above the per-request ceiling. The FAQ says registered users’ synthesis history is stored for 3 days.
Before registering, ask which commercial-use terms apply to the specific voice model you plan to use; the offer requires separate licensing for models it names. Also clarify how token use is calculated for the intended voice, since the FAQ gives different example units for regular, PRO, and PRO+ voices.
Text settings lead to a download
The documented workflow starts with choosing a voice model and synthesis parameters for the text. Where pronunciation stress needs marking, the FAQ says to place a plus sign before the stressed vowel.
After synthesis, users can download the resulting audio as an MP3 or WAV file, according to the provider. Registered users’ synthesis history is stored for 3 days, the FAQ states, giving them a short period to revisit prior outputs.
Documented access covers common audio tasks
The provider lists videos, IVR menus, and automated calls as intended text-to-speech uses.
Users can download the synthesized output as an MP3 or WAV file, according to the homepage.
The offer describes voice-model and synthesis-parameter selection, while the FAQ explains stress marking with a plus sign before the stressed vowel.
The FAQ says multilingual voice models can synthesize text in any language.
The offer grants users exclusive rights to successfully created audio files, subject to the offer’s terms.
Caps and licensing narrow the offer
The provider’s offer and FAQ set conditions around token use, voice rights, and the text submitted for synthesis.
- The homepage says PRO+ voices cannot be paid for with bonus tokens; users must top up their balance instead.
- Commercial use of the voice models named in the offer requires a separate license agreement, although the FAQ says almost all voices may be used commercially.
- Users must hold rights to the text they submit, according to the FAQ.
- The offer says tokens expire 180 calendar days after crediting, and unused tokens are non-refundable except where Russian law expressly provides otherwise.
- The offer does not guarantee that synthesized speech will match users’ subjective expectations of voice characteristics.
- The FAQ warns that mixed Russian-English text may be interpreted incorrectly and spoken with a strong accent.
Several details are missing; they are listed under What we could not confirm.
Fit follows text length and usage rights
Robivox’s stated use cases include videos, IVR menus, and automated calls, so the offer is framed around producing speech for those formats. The fit also depends on whether users can register when longer text is involved and secure rights to both the submitted text and the chosen voice model.
A short audio task may sit within the access conditions, while commercial use of a model named in the offer requires a separate license agreement. Users assessing a project should identify the intended voice model and clarify its applicable rights before committing to that use.
Compare voice tools by rights and limits
When comparing speech tools, check voice and language coverage, audio formats, text limits, commercial rights, charging basis, trial access, and support routes. Ask each provider to state which conditions apply to the chosen voice and intended use.
What we could not confirm
- Payment page: contents could not be read.
- API: synthesis access or integration details.
- Trial: availability and applicable conditions.
- Published tariffs: plan names, current prices, and full allowances.
- Voice catalog: complete list and language coverage by voice.
- Commercial licensing: named models, scope, and agreement terms.
- Token billing: calculation method beyond stated examples.
- Support: response times and service hours.
Questions about Robivox’s terms
Does registration affect text length?
Yes, registration is required for text longer than 100 characters, according to the homepage. The FAQ says unregistered users can synthesize up to 100 characters, while a single request is capped at 5,000 characters, including spaces.
Can bonus tokens pay for PRO+ synthesis?
No, bonus tokens cannot be used with PRO+ voices; the homepage tells users to top up their balance. The homepage also limits PRO+ voices to registered users. The FAQ gives a separate PRO+ example based on characters rather than minutes.
When do tokens expire, and are unused tokens refunded?
The Terms say tokens expire 180 calendar days after they are credited. The Terms also say unused tokens are not refundable except where Russian law expressly provides otherwise. These are the stated expiry and refund conditions; users should check how they apply to the balance they plan to purchase.
What commercial rights apply to named voice models?
The offer requires a separate license agreement for commercial use of the voice models it names. The FAQ separately says almost all voices on the site may be used commercially. Users must also have rights to the text they submit, according to the FAQ.
How might mixed Russian-English text be spoken?
The FAQ warns that text combining Russian and English may be interpreted incorrectly and spoken with a strong accent. This is a provider warning about mixed-language input, not a guarantee of how any particular text will sound. The offer also says subjective expectations of voice characteristics may not be met.
How can users mark stress or contact support?
The FAQ says users can mark stress by placing a plus sign before the stressed vowel. For technical problems, the offer lists a Telegram bot or email as support channels, while the FAQ directs questions to Telegram.
Proceed after resolving the three conditions
Fits if the request length, intended voice model, and text rights all work within the stated terms.
Probably not if commercial use depends on a named model and its separate license terms remain unresolved.
Main risk… token expiry and refund conditions affecting a paid balance, alongside the distinction between general commercial-use wording and named-model licensing.
Sources
Public pages the facts above were taken from, with the date each was read.
- robivox.ruchecked
- robivox.ru/faqchecked
- robivox.ru/ofertachecked
How we put this page together
Editorial team of Affiliate Index. The editors at Affiliate Index wrote this page from the public sources listed above, which were read on 2026-10-31. Nobody tried the service in practice. If a source changes, the page is revised.
Last updated:
Right of reply
Representatives of Robivox can contest this entry: write to us and attach evidence of the inaccuracy. We correct the entry or remove it once the refutation is documented. Entries are never removed or softened for payment.
- Our contact for replies
- No data available
- Response time
- No data available