"RecText AI 2," a Perpetual License AI Transcription Software with Newly Added Speaker Separation Function, to be Released on March 31 (Tues)
SourceNext Corporation is releasing "RecText AI 2," a perpetual license AI transcription software on March 31st. This offline application features a new speaker separation function for improved meeting minute creation and ensures data privacy. It offers fast transcription, batch processing, and enhanced memo capabilities, priced at 9,980 yen for Windows.
📋 Article Processing Timeline
- 📰 Published: March 31, 2026 at 20:00
- 🔍 Collected: April 1, 2026 at 13:39 (17h 39m after Published)
- 🤖 AI Analyzed: April 17, 2026 at 00:46 (371h 6m after Collected)
SourceNext Corporation (Headquarters: 3-8 Sanbancho, Chiyoda-ku, Tokyo, Izumikan Sanbancho 3rd Floor, President: Noriyuki Matsuda) will release "RecText AI 2" (9,980 yen, tax included, download version) on our website on Tuesday, March 31st.
This product is the latest version of a perpetual license AI transcription software that adopts the offline engine of the AI meeting service "AutoMemo," which is provided via subscription. Its offline operation prevents the leakage of confidential information, and the newly added "speaker separation function" automatically identifies speakers, thereby reducing the effort required for creating meeting minutes.
**About "RecText AI 2":**
This product is an AI transcription software that incurs no subscription or usage fees. Because there are no server operating costs, it is offered as a perpetual license rather than a subscription. Despite being a lightweight engine that operates on general-purpose PCs, it features highly accurate transcription with a success rate of 92.1% (Note).
Since no data is transmitted over the internet, it can be introduced by companies that require confidentiality without needing security evaluations.
(Note) Measurement of transcription accuracy: Text generated from a single verification run, transcribing audio from one speaker recorded in a conference room with ambient noise of approximately 40db. (SourceNext research / June 2025). Accuracy varies significantly depending on speaking style and environment.
**Features of "RecText AI 2":**
■ Newly Equipped with Speaker Separation Function *NEW*
It is now instantly clear "who said what" in meetings. Speeches are separated by speaker, and by registering speaker names later, it becomes obvious who spoke. When AI generates summaries, speaker information is included for organization, significantly improving the efficiency of meeting minute creation.
■ Transcription Completed in as Little as 1 Minute After Recording/Video Ends *NEW*
Transcription occurs in the background in parallel with recording and video capture. This reduces the perceived completion time. Transcription is completed approximately 1 to 3 minutes after the recording or video finishes.
(Note) Processing time may be longer depending on PC specifications.
■ Automatic Transcription by Specifying a Folder *NEW*
Automatically transcribes video and audio files within a designated folder. It can process multiple recorded/video data collectively, allowing for efficient transcription of large volumes of files.
■ Batch Transcription of Accumulated Files *Power UP*
By reading pre-recorded files, it is now possible to import multiple files at once. It supports file formats recorded with voice recorders and smartphones. The supported formats have been expanded to m4a, allowing for handling a wider range of files.
Supported formats: mp4, wmv, wav, mp3, m4a
■ Enhanced Memo Function Before/During Recording *Power UP*
During video/audio recording, simply press the bookmark button if you think, "I want to review this later." You can leave the time and, if necessary, comments. Simultaneously, you can also take notes. Since these are saved linked to the video/audio data, your notes won't get lost.
Editing has become smoother with the ability to take memos and set filenames even before recording starts.
**Product Overview**
Product Name, Price: "RecText AI 2" - 9,980 yen (tax included)
Product Content: Offline AI transcription software
Development, Sales, Support: SourceNext Corporation
Product Information: https://www.sourcenext.com/product/rectext/
**Operating Environment**
Supported OS: Windows 11, Windows 10 (64-bit)
CPU: Core i3 (4 cores or more), or equivalent CPU (Recommended: Core i7 or higher)
Memory: 8GB or more (Recommended: 16GB)
Restrictions: Processing time will be slower on 2-core PCs.
**About Copyright Notation**
For copyright notation, such as displaying product screens, please use the following:
© SOURCENEXT CORPORATION
**Customer Inquiry**
SourceNext Customer Center
Pre-purchase Consultation Desk
https://support.sourcenext.com/fa/support/web/form155.html
This product is the latest version of a perpetual license AI transcription software that adopts the offline engine of the AI meeting service "AutoMemo," which is provided via subscription. Its offline operation prevents the leakage of confidential information, and the newly added "speaker separation function" automatically identifies speakers, thereby reducing the effort required for creating meeting minutes.
**About "RecText AI 2":**
This product is an AI transcription software that incurs no subscription or usage fees. Because there are no server operating costs, it is offered as a perpetual license rather than a subscription. Despite being a lightweight engine that operates on general-purpose PCs, it features highly accurate transcription with a success rate of 92.1% (Note).
Since no data is transmitted over the internet, it can be introduced by companies that require confidentiality without needing security evaluations.
(Note) Measurement of transcription accuracy: Text generated from a single verification run, transcribing audio from one speaker recorded in a conference room with ambient noise of approximately 40db. (SourceNext research / June 2025). Accuracy varies significantly depending on speaking style and environment.
**Features of "RecText AI 2":**
■ Newly Equipped with Speaker Separation Function *NEW*
It is now instantly clear "who said what" in meetings. Speeches are separated by speaker, and by registering speaker names later, it becomes obvious who spoke. When AI generates summaries, speaker information is included for organization, significantly improving the efficiency of meeting minute creation.
■ Transcription Completed in as Little as 1 Minute After Recording/Video Ends *NEW*
Transcription occurs in the background in parallel with recording and video capture. This reduces the perceived completion time. Transcription is completed approximately 1 to 3 minutes after the recording or video finishes.
(Note) Processing time may be longer depending on PC specifications.
■ Automatic Transcription by Specifying a Folder *NEW*
Automatically transcribes video and audio files within a designated folder. It can process multiple recorded/video data collectively, allowing for efficient transcription of large volumes of files.
■ Batch Transcription of Accumulated Files *Power UP*
By reading pre-recorded files, it is now possible to import multiple files at once. It supports file formats recorded with voice recorders and smartphones. The supported formats have been expanded to m4a, allowing for handling a wider range of files.
Supported formats: mp4, wmv, wav, mp3, m4a
■ Enhanced Memo Function Before/During Recording *Power UP*
During video/audio recording, simply press the bookmark button if you think, "I want to review this later." You can leave the time and, if necessary, comments. Simultaneously, you can also take notes. Since these are saved linked to the video/audio data, your notes won't get lost.
Editing has become smoother with the ability to take memos and set filenames even before recording starts.
**Product Overview**
Product Name, Price: "RecText AI 2" - 9,980 yen (tax included)
Product Content: Offline AI transcription software
Development, Sales, Support: SourceNext Corporation
Product Information: https://www.sourcenext.com/product/rectext/
**Operating Environment**
Supported OS: Windows 11, Windows 10 (64-bit)
CPU: Core i3 (4 cores or more), or equivalent CPU (Recommended: Core i7 or higher)
Memory: 8GB or more (Recommended: 16GB)
Restrictions: Processing time will be slower on 2-core PCs.
**About Copyright Notation**
For copyright notation, such as displaying product screens, please use the following:
© SOURCENEXT CORPORATION
**Customer Inquiry**
SourceNext Customer Center
Pre-purchase Consultation Desk
https://support.sourcenext.com/fa/support/web/form155.html