Skip to main content
Speech to text

MAI Transcribe 2 speech to text for real conversations

Pick the job you actually have — a meeting, a podcast, captions, or notes — then run a short file. See what the transcript looks like before you commit a longer recording.

Meetings

Turn standups and client calls into searchable notes with speaker labels.

Podcasts

Produce show notes and quotes from long interviews without manual typing.

Captions

Create subtitle drafts for video with timestamps you can clean later.

Doctor’s notes

Dictate follow-up visits and keep medical terms readable.

Run a short clip

Studio

1 free daily · sign in
Drop audio or video here or click to upload
MP3 / WAV / M4A / FLAC / MP4 · video uses the soundtrack · 10MB · 5 minutes

Default. Keep ums, false starts, and spoken wording.

Up to 100 terms. Plus and Pro only.

Comparing models? Compare MAI Transcribe 2 vs Whisper. MAI Transcribe 2.

Files, languages, and privacy

Accepted formats

MP3, WAV, M4A, FLAC, AAC, OGG, and MP4 / MOV / WebM video (audio only). Free: 5 minutes / 10 MB. Plus: 30 minutes / 50 MB. Pro: 60 minutes / 50 MB.

Languages

You do not pick a language first. MAI Transcribe 2 auto-detects among 60 languages and can follow a switch mid-sentence. Still review names.

What happens to the audio

The file passes through our Cloudflare Worker and a contracted speech-processing service. We do not keep a permanent recording. See the privacy policy for retention and deletion.

When a job fails

Rejected file

Wrong type or over the plan limit. Re-export M4A/MP3 or split the recording.

Empty transcript

No text came back. Paid credits are not taken. Try a louder 2-minute slice from the middle.

One speaker for everyone

Common on a single mic in a loud room. Use the draft as notes; do not treat it as a legal who-said-what record.