EKHOS AI is offline Windows transcription in 98 languages with no cloud upload. Free plan available; Premium from $9/month billed annually.
Researched and fact-checked by Bartu Cavusoglu · Updated October 26, 2026
EKHOS AI is a transcription application for Windows that converts audio and video recordings into text on the user's own computer. The company positions it as professional-grade software for legal, healthcare, research, education, law enforcement and media users, as well as anyone concerned about privacy.
Its central claim is that nothing leaves the device. The site states there is no cloud, no telemetry and no usage analytics, and that transcripts and source files are never transmitted externally or used to train its AI models.
EKHOS AI was founded in 2024 by a small start-up based in Australia. The founders say they wanted a transcription service that depends less on the cloud, especially for medical and legal work where recordings are sensitive.
Users sign up with a full name and email address, confirm the email, and install the app from the Microsoft Store. After signing in, transcription runs offline with no internet connection required. The AI model is stored and run locally.
Processing can use the computer's CPU, or an NVIDIA GPU for faster results, with RTX models recommended. Users choose among Intermediate, Advanced and Expert models, and among Optimized, Standard and Speaker Identification modes. The site says Optimized mode is twice as fast as Standard.
Files can be added individually or queued in bulk. The app can also record microphone or speaker audio and transcribe it, starting after one minute of recorded audio by default, a delay that can be configured.
On the transcription side, EKHOS AI handles most audio and video formats, including MP3, MP4, WAV, MKV, AVI and MPEG. It supports 98 languages and offers speaker identification and labeling for recordings with several voices.
For proofreading, a built-in media player lets reviewers play, pause and fast-forward the source file. The Tracks Editor highlights transcript segments in sync with playback so errors can be corrected in place, and a search function lists every track that contains a given phrase. A plain text editor is available as an alternative. The site says these tools help users reach 99% accuracy.
Security options include a local password, recovery phrase, masking and a read-only mode, with transcripts encrypted in a local database. Finished transcripts export to Word, PDF or text, or copy to the clipboard.
EKHOS AI is transcription software for Windows that converts audio and video into text on your own computer. It works offline, supports 98 languages and does not upload files to the cloud.
Premium costs $12 per month billed monthly, or $108 per year billed annually, which is $9 per month. A Free plan costs $0 and allows one transcription of up to 30 minutes per day.
Yes. The Free plan covers one 30-minute transcription per day with microphone and speaker audio, a text editor, text export and support for over 98 languages. It is limited to a single user per computer.
Yes. Offline transcription becomes available immediately after signing in, and no internet access is needed to transcribe. Files and transcripts are processed and stored locally.
The company states there is no telemetry, no usage analytics and no external transmission of files or transcripts. Transcripts are encrypted in a local database and are never used to train its AI models.
It transcribes most audio and video formats, including MP3, MP4, WAV, MKV, AVI and MPEG. It supports 98 languages, including English, German, French, Italian, Spanish, Portuguese, Japanese and Chinese.
It runs on a computer's CPU, and can use an NVIDIA GPU, with RTX models recommended, for faster transcription. The company publishes a hardware requirements page with full details.
The pricing page directs enterprise and bulk subscription inquiries to support@ekhos.ai. No separate enterprise pricing is published.