Add a file, record your voice or generate media for a sequence message. Review the result before saving the workflow. None of these steps sends a message by itself.
You can add media to a LinkedIn message, LinkedIn InMail or email step. Connection requests do not support attachments.
Upload a file or reuse media
Open your agent’s Sequence, choose Open workflow builder and select the message step.
Find Media in Step details.
Choose Upload, then Choose file or drop one file into the upload area. Choose Library instead to reuse saved media.
Wait until the upload finishes and the file preview appears on the step.
Choose Save to apply the workflow. For a new agent, choose Done and finish setup separately.
Check the supported file types and total attachment size. Sample data shown.
Supported files are JPEG, PNG, WebP, GIF, PDF, MP4, M4A, MP3 and WAV. Each step can hold up to four attachments, with 15 MB total. Use a file’s remove button to take it off the step.
Uploading or reusing a file does not use a media-generation post credit. Upload starts share a workspace limit across Inbox and the sequencer, up to 10 starts per 10-minute window. If you reach the limit, the dialog tells you when to try again. Existing files and queued sends are unaffected.
Choose how LinkedIn sends the file
LinkedIn message can send files as attachments. With exactly one audio file, you can choose Send as voice note. With exactly one MP4, you can choose Send as video message.
LinkedIn InMail uses attachments only. It does not support the LinkedIn voice-note or video-message format and does not show Record. InMail also requires eligible Classic Premium InMail credits.
Email sends audio, video and other files as attachments.
For a native LinkedIn voice note or video message, keep only the matching file on the step, then choose its sending format below the preview. Adding another attachment switches the step back to attachment mode.
Record your own voice
Select a LinkedIn message or email step with no attachments.
Choose Record > Start recording and allow microphone access.
Record for up to one minute, then choose Stop recording. The recorder stops automatically at the time limit.
Wait for preparation to finish, then listen to the preview. Use Record again if you want to replace it.
Choose Add recording, then save the workflow.
A recording on a LinkedIn message is added as a voice note. On email, it is an audio attachment. If recording is unavailable, use Upload audio instead. If microphone access was denied, allow it in your browser’s site settings and try again.
Generate an image or AI voice
Select a supported message step with no attachments and choose Media > Generate.
Choose Image or AI voice.
Describe the image, or choose Kore, Puck, Zephyr or Charon and enter the exact script to speak.
Check your available credits, then choose Generate for 1 credit for an image or Generate for 2 credits for AI voice.
Review the result and choose Use image or Use voice note to attach it. Save the workflow when ready.
Review the script, voice and displayed credit cost. Sample data shown.
An image generation uses 1 post credit. AI voice uses 2 post credits. Generate another starts a new generation at the displayed cost. Generating does not send a message, and the result is not attached until you choose Use image or Use voice note.
AI voice accepts up to 800 characters of finished text. Image descriptions can contain up to 2,000 characters. The generated media is reused for every lead who reaches the step. Personalization tokens such as {{firstName}} are not supported in the spoken script.
Only a regular LinkedIn message can use the native voice-note format. AI voice added to InMail or email is an audio attachment. Video generation is not offered here, so upload an MP4 for video.
Review before saving
Finish uploading, recording or generating before changing steps or saving. Check the file preview, sending format and message text. The sequence walkthrough shows that media is included but does not send it.
Pause an active agent before major workflow changes and review the sequence before resuming. Approvals, sending hours, account limits and channel requirements still apply.