Voice input: Start recording – User guide

Capture and transcribe voice messages directly in Chat. Instead of typing, speak your message and let AI transcribe it automatically.

Use voice input when you prefer speaking over typing, or when you want to quickly dictate messages while multitasking.

Overview

The Start recording feature lets you record audio messages in Chat, which are automatically transcribed into text. After transcription, the text appears in your input field where you can review, edit, and send it. This feature supports multiple languages, making it accessible for users worldwide.

Voice input is ideal for quick dictation, hands-free interaction, or when you want to express ideas naturally through speech.

Prerequisites

To use voice input, you need:

• A Swiftask account (sign up at swiftask.ai)

• Access to Chat

• A microphone on your device

• Microphone permissions enabled in your browser (the system will prompt you on first use)

Step-by-step guide

1. Open Chat and locate the microphone icon

Navigate to Chat. In the input area at the bottom of the screen, look for the toolbar icons on the right side of the input field.

The toolbar contains several icons for different actions. The microphone icon (shown in the image below) is located in the right toolbar, between the magic wand icon (Enhance with AI) and the wave icon (Start conversation).

2. Click the microphone icon to start recording

Click the microphone icon to begin recording your voice message.

The recording interface opens immediately, showing:

• Timer – Displays elapsed time (e.g., 00:03)

• Waveform display – Visual representation of your audio

• Cancel button (X) – Stop recording and discard

• Confirm button (✓) – Stop recording and transcribe

As shown in the image below, the recording interface is ready for your voice input.

3. Speak your message clearly

Speak your message naturally into your microphone. The waveform display updates in real time as you speak, showing that your audio is being captured.

Keep speaking until you've completed your message.

4. Confirm your recording

Once you finish speaking, click the checkmark button (✓) to confirm and transcribe your recording.

A confirmation dialog appears, as shown below, indicating that your audio is being processed.

5. Review and edit the transcribed text

After transcription completes, the transcribed text automatically appears in your Chat input field. You can now:

• Review the transcribed text for accuracy

• Edit the text if needed (correct any words or add context)

• Send the message by pressing Enter or clicking the send button

As shown in the image below, the transcribed message appears in the input field, ready for you to send or modify.

6. Send your message

Press Enter or click the send button (up arrow icon) to send your transcribed message to Chat.

Chat will respond to your message as it would to any typed message.

Supported languages

Voice input supports transcription in multiple languages, including:

• English

• And many other languages

Speak in your preferred language, and the transcription will adapt automatically.

Practical use cases

Quick brainstorming: "Record a voice message asking for help brainstorming marketing campaign ideas for a new product launch." Speak your thoughts naturally, and Chat will transcribe them instantly.

Hands-free messaging: While driving or multitasking, dictate a message instead of typing. Click the microphone icon, speak, confirm, and send.

Non-English speakers: Dictate messages in your native language. Voice input transcribes and Chat responds in the language you use.

Capturing ideas on the go: Use your phone or laptop microphone to quickly capture an idea before you forget it. No need to type it out manually.

Tips & best practices

Speak clearly and at a natural pace – The transcription works best when you speak clearly without rushing. Use natural pauses between sentences.

Use punctuation in your speech – If you need specific punctuation, you can say "period," "comma," "question mark," or "exclamation point" while recording, or add it manually when editing.

Edit before sending – Always review the transcribed text before sending. Transcription is accurate but may occasionally need minor corrections, especially with technical terms or names.

Keep messages concise – While there's no strict time limit, shorter recordings are easier to transcribe accurately. If you have a long message, consider breaking it into two recordings.

Cancel and retry if needed – If you make a mistake while recording, click the X button to cancel and start over. There's no penalty for retrying.

Troubleshooting

Microphone not detected

Cause: Your device doesn't have a microphone, or it's not connected.

Solution:

• Check that your microphone is properly connected (for external microphones)

• Ensure your device's built-in microphone is functional

• Try using a different microphone if available

Transcription is inaccurate

Cause: Background noise, unclear speech, or unfamiliar words.

Solution:

• Speak more clearly and at a natural pace

• Record in a quiet environment to reduce background noise

• Manually edit the transcribed text in the input field before sending

• For technical terms or names, spell them out or add them manually after transcription

Recording won't start

Cause: Microphone permissions not granted to Swiftask.

Solution:

• Check your browser settings and allow microphone access for swiftask.ai

• Refresh the page and try again

• Clear your browser cache and cookies, then retry

Additional resources

• Chat – Getting started – Learn how to start your first conversation in Chat

• Chat – Attaching files & documents – Combine voice input with file analysis for richer interactions

• Browse and change AI models – Choose the best AI model for your voice-based conversations

Ready to use voice input? Click the microphone icon in Chat now and start dictating your messages!