> For the complete documentation index, see [llms.txt](https://docs.maiagent.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.maiagent.ai/maiagent-user-guide/en/work/meeting-records/offline-transcription.md).
# Offline Meeting Transcription
Offline meeting records: upload audio files for multi-language high-accuracy transcription, speaker identification and management, timeline transcripts, dynamic subtitle playback, and supported audio
Offline mode lets you upload pre-recorded audio files. The system performs high-accuracy transcription and speaker identification, and generates a transcript with a timeline.
## Supported Audio Formats
| Format | Extension |
| ------ | --------- |
| WAV | `.wav` |
| MP3 | `.mp3` |
| M4A | `.m4a` |
| AAC | `.aac` |
The file picker lists only these four formats. For video files (such as `.mp4`), convert them to one of the audio formats above before uploading.
## Upload Audio
1. Go to the Meeting Records page and click Upload Offline Audio in the upper-right corner (transcription permission is required; otherwise the button is grayed out with the message "You do not have permission to upload audio files for transcription").
2. Fill in the following under "Transcription Settings":
| Field | Description |
| --------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Meeting Name | Required. You can [change it](/maiagent-user-guide/en/work/meeting-records/meeting-detail.md#edit-meeting-name) later on the detail page |
| Offline Transcription Provider | Required. The list depends on the providers enabled for your organization. If "No offline transcription providers are currently available" is displayed, contact your administrator |
| Language | Required. Multiple selections are allowed; select every language that will be spoken in the meeting for better recognition accuracy |
| Speaker Identification | When enabled, the completed transcript distinguishes between different speakers; when disabled, the system determines the audio tracks automatically. If the selected provider does not support it, the message "The selected ASR model does not support speaker identification" is displayed and the switch is unavailable |
| Provider advanced options | Some providers offer additional parameters (marked "Optional"). If you are unsure, leave them blank |
3. In the Audio File section, click or drag your file into it ("Click or drag an audio file here to select"). If the format is not supported, the message "This audio file format is not supported" is displayed.
4. Click Start Transcription. The screen first displays "Uploading audio file", and after the upload completes it shows "Audio file uploaded. Transcription in progress."
After the upload, the meeting record status changes to "Processing" and the system transcribes in the background; you can leave this page and work on other tasks. If transcription fails, the detail page displays "Audio transcription could not be completed. Please upload again or try again later." Simply upload the file again.
## Multi-language High-accuracy Transcription
Offline mode uses a high-accuracy speech recognition engine (such as WhisperX), which provides higher recognition accuracy than live mode:
* Supports multi-language recognition
* Higher recognition accuracy than live transcription
* Ideal for important meetings that require precise records
## Speaker Identification and Management
Offline mode supports automatic Speaker Diarization, in which the system automatically distinguishes between different speakers:
* **Auto Identification**: The system automatically labels different speakers (e.g., SPEAKER\_00, SPEAKER\_01)
* **Custom Names**: Replace system labels with real names (e.g., "Manager Zhang", "Team Lead Li")
* **Management Interface**: Manage the label names of all speakers in one place
After you change a speaker label, all related segments in the transcript are updated accordingly.
## High-accuracy Timeline Recognition
Each transcribed text segment includes precise start and end times, so you can:
* Pinpoint the exact time of each utterance
* Jump in sync with the audio player
* Export as SRT subtitle files with timestamps
## Dynamic Subtitle Playback
Offline mode provides dynamic subtitles that sync the audio player with the transcript:
* **Sync Highlighting**: During audio playback, the transcript segment currently being played is automatically highlighted
* **Click to Jump**: Click any transcript segment, and the audio automatically jumps to the corresponding time point
* **Search and Navigate**: After searching for a keyword, you can jump directly to the audio position of that segment
This feature lets you quickly find specific discussions in a meeting without listening to the entire recording from the beginning.
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.
Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:
```
GET https://docs.maiagent.ai/maiagent-user-guide/en/work/meeting-records/offline-transcription.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.