The subject does not meet Wikipedia's criteria for inclusion. Rejected by Bonadea 60 days ago. Last edited by Bonadea 60 days ago. |
Submission declined on 27 February 2026 by Theroadislong (talk). Declined by Theroadislong 3 months ago. |
This draft appears to contain text generated by a large language model (such as ChatGPT). You cannot use LLMs to generate article content.
Declined by Nil NZ 3 months ago.LLM-generated pages with certain obvious signs of being machine generated may be deleted without notice. These tools are prone to specific issues that violate our policies:
Instead, only summarize in your own words a range of independent, reliable, published sources that discuss the subject. See the advice page on large language models for more information. |
Submission declined on 25 February 2026 by Theroadislong (talk). This draft's references do not show that the subject meets Wikipedia's criteria for inclusion. The draft requires multiple published secondary sources that:
Declined by Theroadislong 3 months ago.
|
Comment: No independent sources, no claim to notability. bonadea contributions talk 19:37, 8 April 2026 (UTC)
Comment: GPTZero says highly confident this text was 100% AI generated. Theroadislong (talk) 07:57, 27 February 2026 (UTC)
Comment: There are some statements that appear hallucinated, or are not entirely supported by the cited source. For example: none of the sources say "33 languages"; source [1] doesn't mention anything about ten simultaneous speakers...
; & source [3] doesn't mention MP4's etc.Other likely AI tells include the promotional tone, list formatting (instead of prose), and random mark-up. Please resubmit the draft, without the use of an LLM. nil nz 05:19, 27 February 2026 (UTC)
Comment: Please stop removing the decline templates. These are preserved to make it easier for future reviewers to review your draft and check that any prior concerns have been properly addressed. Removing this template can make it more difficult for us to accept your draft. guninvalid (talk) 05:05, 27 February 2026 (UTC)
| Perso AI Dubbing | |
|---|---|
| Developer | ESTsoft |
| Initial release | May 2024 |
| Operating system | SaaS (Web-based) |
| Available in | Multilingual (33+ languages) |
| Type | Video localization, Speech synthesis |
| License | Proprietary |
| Website | perso |
Perso AI Dubbing is a web-based video dubbing platform developed by ESTsoft, a South Korean software company. The service processes video files and regenerates the audio track in a target language while preserving the original speaker's vocal characteristics.[1] Launched in May 2024, the platform supported more than 33 languages as of early 2026.[2]
History
editESTsoft launched Perso AI Dubbing in May 2024, initially with support for 29 languages.[3] A voice cloning feature was introduced in October 2024, enabling the dubbed audio to more closely replicate the timbre of the original speaker.[4] The following month, ESTsoft added a script editor allowing users to review and revise machine-translated text before the final dubbed audio is rendered.[5]
In 2025, ESTsoft integrated the ElevenLabs voice synthesis engine into the platform.[6] The platform was selected for the South Korean government's K-FAST programme, which supports AI dubbing of Korean media for international distribution.[7] By September 2025, the platform had reached 100,000 registered users.[8]
In October 2025, the service recorded notable growth in Russia, Kazakhstan, and Türkiye, with prominent Russian YouTube creators among its users.[1] The following month, ESTsoft introduced LLM-based dubbing, a technique designed to improve sentence-length alignment and timing synchronisation across languages.[9] Also in late 2025, AI-dubbed K-content began streaming on LG and Samsung TV Plus channels through the FAST distribution model.[7]
At CES 2026, ESTsoft demonstrated an "AI Promoter" system at the Samsung Electronics exhibition that used Perso AI technology to hold real-time multilingual conversations with booth visitors in 32 languages.[10]
Features
editThe platform accepts MP4, MOV, and WebM video files, as well as URLs from YouTube, TikTok, and Google Drive.[2] It can identify up to ten distinct speakers within a single video and outputs a lip-synced MP4 file alongside an SRT subtitle file.[2]
Core capabilities include AI dubbing across 33+ languages, lip-sync adjustment to match new audio to mouth movements, voice cloning to approximate the original speaker's tone, and a script editor for manual correction of translated text before audio generation.[4][6][5]
Recognition
editThe International Telecommunication Union (ITU) featured Perso AI Dubbing as an innovation case study at its "AI for Good Global Summit".[8] The platform is listed on the Microsoft Azure Marketplace.[2]
See also
editReferences
edit- 1 2 Jo, Yoon-joo (2025-10-20). "ESTsoft Expands Perso AI Dubbing Service into Eurasia". Financial News. Retrieved 2026-04-08.
- 1 2 3 4 "ESTsoft Launches AI Dubbing Platform PERSO.ai Globally". Chosun Biz. 2024-10-25. Retrieved 2026-04-08.
- ↑ "ESTsoft Develops AI Dubbing Service Supporting 29 Languages That Automatically Dubs Videos". Chosun Biz. 2024-09-03. Retrieved 2026-04-08.
- 1 2 "ESTsoft's PERSO.ai Enhances AI Dubbing with Voice Cloning Feature". Chosun Biz. 2024-10-11. Retrieved 2026-04-08.
- 1 2 "ESTsoft's PERSO.ai Launches Script Editing Feature for AI Dubbing". Chosun Biz. 2024-11-08. Retrieved 2026-04-08.
- 1 2 "Dubbing made simpler, sharper, and faster at PERSO.ai". ElevenLabs. Retrieved 2026-04-08.
- 1 2 "ESTsoft Unveils K-FAST AI Technology That Improves Dubbing Quality and Speed". Chosun Biz. 2025-10-24. Retrieved 2026-04-08.
- 1 2 "ESTsoft's AI Dubbing Platform PERSO.ai Surpasses 100,000 Users". Chosun Biz. 2025-09-12. Retrieved 2026-04-08.
- ↑ "ESTsoft presents LLM-based AI dubbing that improves sync and multilingual quality". Chosun Biz. 2025-11-17. Retrieved 2026-04-08.
- ↑ "ESTsoft and Samsung Electronics Unveil 'AI Promoter' at CES 2026". The Korea Herald. 2026-01-15. Retrieved 2026-04-08.
Category:Video software Category:Speech synthesis Category:South Korean websites Category:2024 software


- provide significant coverage: discuss the subject in detail, excluding routine coverage like product launches, staff appointments, or financial reports and listings in databases or listicles;
- are reliable: from reputable outlets with editorial oversight;
- are independent: not connected to the subject, such as press releases, the subject's own website, or sponsored content.
Please add references that meet all three of these criteria. If none exist, the subject is not yet suitable for Wikipedia.