What to expect from a Macedonian draft
Macedonian has far less training data behind it than English, German or even Bulgarian, so treat the output as a draft that needs a full review. Common words and clear sentences usually come through. Names, places, dialect words and technical terms need checking every time.
The best results come from one speaker close to the microphone in a quiet room: a lecture, a dictated report, an interview recorded with a lapel mic. Busy family gatherings, phone calls on speaker, several people talking at once, or speech over music give much weaker drafts. In those cases the timestamps are mostly useful for finding your place while you type the correct text.
Correcting a draft is still quicker than typing Macedonian Cyrillic from scratch, especially if you are on a keyboard set up for Latin letters. Play the file, follow the highlighted line, and fix errors as you hear them. For anything you plan to publish or submit, such as a thesis quote, a court document or a news story, ask a native speaker to read the final text, because a single wrong letter can change a word's meaning or make it read as Serbian or Bulgarian.
Macedonian, Bulgarian, Serbian and auto-detect
Macedonian is close to Bulgarian, and many words are shared with Serbian too. If you leave the language on auto-detect, the model can label Macedonian speech as Bulgarian or Serbian. The transcript then uses the wrong spelling rules: Bulgarian letters such as ъ, щ and я, or Serbian ћ and ђ, where Macedonian would use ќ and ѓ. Short voice notes and recordings that open with a few English words are the most likely to be misjudged.
This page preselects Macedonian to avoid that. If a transcript reads like Bulgarian or Serbian, check the language setting and run the file again with Macedonian selected.
Macedonian Cyrillic: points to check
The transcript is written in the Macedonian Cyrillic alphabet. These are the spots where a review usually finds errors.
- The letters ѓ and ќ can be confused with each other, with г and к, or with Serbian ђ and ћ.
- ѕ (dz), џ, љ and њ may be replaced with two-letter combinations or with similar-looking letters.
- Postpositive articles (-от, -ов, -он and their feminine and plural forms) may be dropped or attached to the wrong word.
- Dialect speech from the east (Štip, Strumica) or the southwest (Ohrid, Struga) may be written closer to the standard language than it was spoken, or misheard.
- Names of people and villages often come out in a common-word spelling, so check each one.
Mixing, the diaspora and common recordings
Many Macedonian speakers mix in English, Serbian or German words, depending on where they live and work. Choose the language by what dominates the file. If a call is mostly English with a few Macedonian sentences, choose English for that file.
Macedonian communities in Australia, Germany, Switzerland, Italy, the US and Canada upload a lot of family material: grandparents talking about village life, wedding videos, phone calls with relatives back home. Other common uploads are university lectures, journalist interviews, research interviews, church talks and YouTube videos that need Macedonian subtitles.
Uploading and what to do next
WhatsApp voice notes, phone recordings and MP4 or MOV videos upload as they are, with no conversion step. The transcript is split into paragraphs with timestamps, and the player highlights the current line. Click a line to hear it again and fix the text in place.
Exports to TXT, DOCX, PDF, SRT, VTT and CSV keep the Cyrillic text as written. To read the content in English, use the Google Translate link on the transcript, or the AI prompt export to ask ChatGPT, Claude or Gemini for a translation or summary. Speaker labels are available on paid plans or with a Minute Pack. Your files are never used to train AI models.