Dolphin Voice TTS Online

What is Dolphin Voice TTS Online?

Dolphin Voice TTS Online is a web-based AI text-to-speech and audio creation platform.

It offers real and digital voice tones, and integrates tools such as multi-person voice recording, audiobooks, subtitles, speech recognition, and AI-generated artwork.

Current service status

The official payment page indicates that some AI voice generation services have been suspended due to compliance requirements related to deep synthesis and generative AI.

The platform states that it is expected to resume operations in the near future; before purchasing a membership or a pack of characters, one should first verify that the functions needed are available and can be used properly.

Main functions

• Convert text into natural speech.
• More than 160 real human voice options are available.
• It offers 2D vocal tones tailored for anime and game content.
• Create voiceovers for dialogues with multiple characters and multiple paragraphs.
• The task is to convert long texts into audiobooks.
• Generate subtitles and provide multilingual translation.
• Identify text from audio.
• Generate a summary for the meeting audio.
• Engage in voice modulation, cover singing, and voice-related creative work.
• Generate images, depth maps, or perform background processing.

Text to speech

After entering the text, selecting a voice tone and adjusting the parameters, the user can generate audio suitable for videos, courses, or narration.

On the website interface, some editing buttons are marked as not yet available; it is not possible to determine whether a function is usable just by looking at its name.

Real-person and 2D voice tones

The current title on the official website highlights the presence of over 160 real human voices, and it provides a separate section for anime-style voices.

• The real-person voice tone is suitable for narration, courses, promotions, and audio content.
• 2D timbres are suitable for character portrayal and fan creation.
• Different timbres can vary in terms of language, emotion, and clarity.
• The voice tones of popular figures or game characters do not automatically receive official authorization.
• For commercial projects, the licensing for sound tones and their scope of use must be verified on a case-by-case basis.

When the timbre sounds like that of a real person or character, it should not be used for impersonation, deception, or unauthorized commercial purposes.

Multiple voiceovers

Multiple voiceovers allow the script to be divided into several segments, with each segment enabling different voice tones, speaking speeds, volumes, and pauses to be set.

1. First, organize the lines by character and remove irrelevant markers.
2. Choose different voice tones for the narration and each character.
3. Adjust the speaking speed and volume for each paragraph to avoid excessive differences between characters.
4. Set pauses between paragraphs to mimic the rhythm of natural conversation.
5. Generate short segments to check whether the characters and lines match.
6. After completing the entire video, handle the background music and subtitles together.

Multiple voiceovers are suitable for radio dramas, short video narratives, course dialogues, customer service demonstrations, and game prototypes.

Audiobook synthesis

The audiobook feature is designed for longer text content, and the character package options also explicitly indicate support for audio playback.

• Split the text into chapters and save them with a consistent naming convention.
• Create a list of pronunciations for characters, narrators, and proper nouns.
• First, determine the overall tone and speaking pace of the entire book.
• Manual proofreading is performed on numbers, English text, and characters with multiple pronunciations.
• Generate chapter by chapter to prevent a failure in one task from affecting the entire content.
• Confirm the book’s copyright and the authorization for an audio adaptation.

Subtitle generation, translation, and speech recognition

Both subtitles and summaries are generated automatically by AI, but they require manual review before they are officially released or before meeting decisions are implemented.

AI drawing and other tools

The navigation system also includes features such as AI-generated graphics, depth maps, face swapping, background replacement, and voice transformation for singing.

• The AI drawing quota varies depending on the membership plan.
• Depth maps can assist in certain visual and 3D creation tasks.
• Background replacement involves the rights to the characters and elements in the original image.
• Face swapping must not be used for impersonation, fraud, or pornographic purposes.
• For cover versions with voice transformation, it is necessary to verify the song, the accompaniment, and the licensing for the voice usage.
• The current status of various functions is subject to the actual page.

Tutorial for dubbing dolphins

1. Visit the official website and first check the announcements regarding service suspensions or resumptions.
2. Choose between normal voiceover, multiple voices, or audiobook tasks.
3. Paste the text and remove any private or unauthorized content.
4. Choose the language, as well as the real-person or anime voice style.
5. Set the speech rate, volume, pauses, and paragraph roles.
6. Generate a trial playback using short text and check the pronunciation.
7. Correct polyphonic characters, numbers, English words, and proper nouns.
8. Submit the complete task only after confirming that the function is working properly.
9. Review the timeline and translation when generating subtitles.
10. Download the audio, subtitles, and any necessary engineering materials.
11. After payment, save the exclusive member link containing the token.
12. Check the commercial licensing before publishing and add an AI-generated identifier.

Member price

The information on the official website regarding current packages was verified as of August 31, 2026; prices, original prices, and availability may change.

“Permanent” generally refers to the account benefits available during the product’s lifespan; it does not mean that the services, models, or all functions will remain unchanged forever.

Comparison of member benefits

For the subtitle tasks under annual and permanent plans, the maximum duration per task is currently set at 1.5 hours, with a maximum upload size of 500 MB.

Price per character pack

Character packs are more suitable for users who have high usage volumes but do not require a daily quota over the long term, and they support cumulative top-ups.

All four character packs indicate support for audio playback; it is still necessary to check the remaining quota based on the actual tasks before they expire.

How to choose between members and character packs

How are characters calculated?

According to the official guidelines, characters in Chinese, English, spaces, and symbols are all counted as individual characters, with each one being considered one character.

• Punctuation, spaces, and line breaks can also consume quota.
• The character quota for members is automatically refreshed at midnight every day.
• Member characters that are not used on a given day are not carried over to the next day.
• Character packages are used based on their total amount and validity period, rather than as a daily membership quota.
• Regenerating and modifying it, then trying again, may also result in additional charges.

Exclusive link account mechanism

Platform members do not need to register for a traditional account; after making a payment, a unique link containing a token is generated for them.

1. Do not close the page or navigate away manually after the payment is successful.
2. Wait for the system to automatically display and open the dedicated link.
3. Save the link immediately in a secure password management tool.
4. Do not expose links in group chats, public documents, or screenshots.
5. If the link is lost, use the merchant’s order number to search or contact the authorities.
6. Upon discovering that a link has been leaked, consult the customer service promptly to find out how to deal with it.

Anyone who obtains the exclusive link can share the membership benefits, so it should be protected just like a password.

Boundaries between commercial use and copyright

• Original copywriting is usually more suitable for commercial dubbing.
• To read aloud or adapt other people’s articles, novels, and scripts, permission is required.
• The voice of 2D characters may be related to the rights regarding the work, the characters, and their performance.
• Voice tones that resemble those of real people must not be used for fraudulent endorsement or identity deception.
• When covering a song, it is necessary to check for permissions regarding the lyrics, music, recording, and backing track.
• For face swapping and using reference images, permission for the use of portraits and pictures is required.
• Comply with the requirements for labeling AI-generated content when releasing it to the public.

Privacy and data security

1. Do not submit ID cards, account passwords, or confidential client information.
2. Obtain the participants’ consent before uploading the meeting audio.
3. Voice replication uses only the individual’s own samples or samples that have been explicitly authorized.
4. Do not share the exclusive member link with unauthorized persons.
5. After the download is complete, remove any cloud tasks that are no longer needed.
6. For sensitive corporate projects, storage and deletion rules must be established first.
7. Keep backups of your original files and the final products.

API and open-source status

The fact that third-party plugins can connect to services does not mean that official use of generic APIs is permitted; authorization must be obtained before integrating them into production systems.

Usage restrictions

• Some AI voice generation services are currently on hold.
• Some editing buttons still show as unavailable.
• Members’ daily character quota does not carry over, which may lead to unused characters.
• Subtitles and recognition can be affected by noise, accents, and multiple speakers.
• There are risks of licensing issues and misguidance associated with similar vocal tones between real people and characters.
• The exposure of exclusive links can lead to others sharing the member benefits.
• The permanent solution still faces product and policy changes.

Frequently Asked Questions

What is TTS Online, the dolphin voice synthesis service, mainly used for?

It offers text-to-speech conversion, real-person and anime-style voices, multi-person voice recording, audiobooks, subtitle translation, speech recognition, and AI-generated graphics.

Can dolphin voiceovers be generated properly now?

The current announcement on the official payment page states that the AI voice generation service is temporarily unavailable; it is advisable to conduct a practical test using the desired voice tone and features before making a purchase.

How much does it cost to become a member for dolphin voiceovers?

On the official website, the price is 25 yuan per month, 65 yuan per quarter, 259 yuan per year, and 499 yuan for a permanent license; the actual prices and availability are indicated on the settlement page.

How much is the dolphin voice pack?

Currently, the cost is 9.9 yuan for 30,000 characters, 19 yuan for 100,000 characters, 69 yuan for 500,000 characters, and 155 yuan for 1 million characters.

Are the characters unused by a member on that day carried over?

No, the character quota for members is reset every day at midnight, and any unused characters are not carried over; character packages, on the other hand, are calculated based on their total amount and validity period.

Why is it necessary to save a special link after making a payment?

The platform uses a unique link containing a token as the membership credential; those who obtain this link can share the associated benefits, so it should be kept securely, just like a password.

Is the dolphin voice synthesis open source?

No, as of now there is no official open-source repository or public API available; the ESP-AI plugins found on the Internet are third-party adaptations.