# Edict [Health: Active]

**Category:** 🎙️ Speech-to-Text  
**Repository:** https://edict.sh/docs  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/edict

## Description
Search your dictation transcripts and transcribe audio or video on-device, with speaker labels.

## Claude Desktop Quick Installation
No confirmed install details yet, so no config is published. Follow the setup steps in the repository above.

## Documentation

## What Edict MCP server does

Edict MCP server exposes two documented capabilities for working with spoken content: searching dictation transcripts and transcribing audio or video. Transcription runs on-device, and the resulting text can include speaker labels. Together, these capabilities support both retrieval from an existing transcript collection and creation of new transcripts from recorded media.

The server fits the Speech-to-Text category because its primary media operation converts speech in audio or video into text. Its transcript search capability is useful after transcription has already taken place, when an agent needs to locate a phrase, topic, or relevant passage in stored dictation. The available description does not specify the supported media formats, search syntax, transcript storage location, language coverage, or speaker-identification method.

## How it works

An MCP-compatible client can use Edict to work with dictation transcripts and request transcription for audio or video. For a new recording, the on-device workflow produces text and may identify different speakers with labels. For existing material, transcript search helps the client retrieve matching content rather than requiring the full transcript to be read manually.

The on-device detail is important for deployments where recordings should be processed locally. Those details should be verified in the Edict documentation before deployment.

## Tools and capabilities

The Edict MCP server is described as supporting:

- Searching dictation transcripts.
- Transcribing audio on-device.
- Transcribing video on-device.
- Adding speaker labels to transcription output.

These capabilities make it a candidate for meeting-recording review, voice-note retrieval, and workflows that need searchable text from recorded speech. The listing does not provide formal tool names, input schemas, output schemas, or examples, so client configuration should not assume particular parameter names.

## Limitations and notes

It does not identify installation commands, required environment variables, supported MCP clients, license terms, operating-system requirements, audio or video formats, languages, or accuracy characteristics. It also does not explain whether speaker labels are available for every input or require particular recording conditions.

Before adopting the Edict MCP server, confirm those operational details in the documentation.

_Full upstream README: https://allmcps.com/mcp/edict/readme_