Advanced Speech to Text Extension for Raycast
A powerful and easy-to-use extension for Raycast that converts voice to text using OpenAI's most advanced models, including GPT-4o-transcribe.
โจ Key Features
- Instant Recording: Starts recording immediately when opening the extension
- Automatic Transcription: Automatically transcribes when you stop recording
- Smart Formatting: Converts your text to different formats (Email, Slack, Report)
- Customizable Prompts: Configure custom prompts for each type of format
- Multiple Models: Support for Whisper-1 and GPT-4o Transcribe
- 99+ Languages: Auto-detection or manual language selection
- History: Save and manage your recording history
๐ Available Commands
Dictate
- Function: Instant recording with automatic transcription
- Usage: Opens directly in recording mode, press Enter to stop and transcribe
Recording History
- Function: View and manage recording history
- Usage: Search, copy or delete previous transcriptions
โ๏ธ Configuration
Basic Configuration
- OpenAI API Key: Your OpenAI API key (required)
- Model: Select between Whisper-1 or GPT-4o Transcribe (Recommended)
- Language: Auto-detection or manual selection
- Temperature: Controls the creativity of transcription
- Response Format: Text, JSON, SRT or VTT
๐ฏ Customizable Prompts (NEW)
You can now fully customize the prompts for each type of format:
Custom Email Prompt
- Purpose: Define how to format transcriptions as professional emails
- Default prompt:
Transform the following text into a professional and well-structured email.
Add an appropriate greeting, organize the information in clear paragraphs, and end with a professional closing.
Maintain a formal but friendly tone. Only return the email text, without subject.
Custom Slack Prompt
Custom Report Prompt
-
Purpose: Define how to format transcriptions as structured reports
-
Default prompt:
Transform the following text into a structured report for a task management system.
Organize the information with:
- **Objective/Task:** [clear description of what needs to be done]
- **Details:** [specific information and steps if any]
- **Requirements:** [if applicable, what is needed to complete the task]
Maintain a clear, professional and action-oriented format.
๐ Custom Prompt Examples
For More Creative Emails:
Convert the following text into a creative and persuasive email.
Use a friendly but professional tone, include an eye-catching introduction
and end with a clear call to action. Organize the information in a way
that is easy to read and convincing.
For Slack with Specific Style:
Transform this text into a Slack message for a development team.
Use appropriate technical terminology, maintain a direct and collaborative tone.
Include relevant development emojis (๐ โ
๐ง etc.) and structure
the information to facilitate quick responses.
For Project Reports:
Convert this text into a project status report with the following sections:
- **Executive Summary:** [brief overview of status]
- **Current Progress:** [what has been completed]
- **Next Steps:** [planned actions]
- **Blockers:** [if any]
- **Delivery Date:** [estimation if relevant]
๐ฎ Workflow
- Open "Dictate" - Recording starts automatically
- Speak clearly - Say your message
- Press Enter to stop and transcribe, or select a specific format (Email, Slack, Report)
- Result handling:
- If using default format: Text is automatically pasted into the currently selected text field
- If using specific format (Email/Slack/Report): Review the formatted result and press Enter to paste
๐ Setup & Installation
Prerequisites
Before installing the extension, you need to set up the following requirements:
1. Install SoX (Audio Processing Library)
This extension requires SoX for audio recording functionality. Choose one of these installation methods:
Option A: Using Homebrew (Recommended)
brew install sox
Option B: Using MacPorts
sudo port install sox
Option C: Download from Official Website
Verify Installation:
sox --version
2. Get Your OpenAI API Key
๐ Privacy Note: Your API key never leaves your computer. All communication is direct between your device and OpenAI's servers.
-
Create OpenAI Account
-
Generate API Key
- Navigate to API Keys section
- Click "Create new secret key"
- Copy the generated key (you won't see it again!)
-
Add Credits (if needed)
- Go to Billing
- Add payment method and credits as needed
Installation Steps
-
Install Extension
- Open Raycast
- Search for "Store"
- Install "Advanced Speech to Text"
-
Configure API Key
- Open Raycast Preferences
- Navigate to Extensions โ Advanced Speech to Text
- Paste your OpenAI API Key in the "OpenAI API Key" field
-
Customize Settings (Optional)
- Select your preferred Whisper model
- Choose default language or keep auto-detection
- Adjust temperature and response format as needed
-
Test the Extension
- Open Raycast and search "Dictate"
- Grant microphone permissions when prompted
- Start your first recording!
๐จ Custom Prompt Configuration
You can fully customize how your transcriptions are formatted for different use cases:
Accessing Custom Prompts
- Open Raycast Preferences
- Go to Extensions โ Speech to Text
- Scroll down to find these customizable prompt fields:
- Custom Email Prompt
- Custom Slack Prompt
- Custom Report Prompt
How to Customize Prompts
Example: Professional Email Style
Transform the following text into a formal business email.
Use professional language, include proper salutations,
organize content in clear paragraphs, and end with
appropriate closing remarks. Maintain executive-level tone.
Example: Creative Slack Messages
Convert this text into an engaging Slack message for a marketing team.
Use relevant emojis, maintain enthusiastic but professional tone,
include call-to-action elements, and format for easy readability.
Add appropriate marketing terminology where relevant.
Example: Technical Documentation
Transform this into technical documentation format with:
- **Overview:** Brief summary
- **Implementation:** Detailed steps
- **Requirements:** Technical dependencies
- **Notes:** Important considerations
Use developer-friendly language and markdown formatting.
Prompt Customization Tips
- Be Specific: Include exact formatting requirements
- Set Tone: Specify formal, casual, technical, creative, etc.
- Include Examples: Reference specific terminology or style
- Define Structure: Outline how information should be organized
- Test & Iterate: Try different prompts to find what works best
๐ง Requirements
- macOS: Compatible macOS version
- SoX: Audio processing library (see installation above)
- Raycast: Latest version recommended
- OpenAI API Key: Active account with available credits
- Microphone: Working microphone with system permissions
- Internet Connection: For OpenAI API communication
๐ก Usage Tips
- For better results: Speak clearly in a low-noise environment
- Languages: Auto-detection works very well, but you can specify the language for greater accuracy
- Custom prompts: Experiment with different prompts to adapt the format to your specific needs
- History: Use the "Recording History" command to review and reuse previous transcriptions
๐ Troubleshooting
- API Error: Verify that your OpenAI key is valid and has credits
- Not recording: Check microphone permissions in System Settings
- Incorrect transcription: Try with specific language instead of auto-detection
- Unwanted format: Adjust custom prompts in preferences
๐ Support
If you have problems or suggestions, you can:
- Review the configuration documentation in Raycast
- Verify that your OpenAI API key is configured correctly
- Try different model and temperature settings
Enjoy using this extension to convert your voice to text efficiently! ๐คโจ