Media library

Sidebar: Content → Media library. Voice messages, videos and documents you can drop into any automation. Record once, reuse everywhere.

Paid plans only. On Free the page explains that media is on the paid plans and links to the plan picker. Uploading, renaming and deleting all need the manage_automations capability — owner, admin and agent; a client can still browse, search and play everything, with a note explaining the read-only state.

The page keeps those two absences separate on purpose: a plan limit is something you can act on by upgrading; a role limit means asking an owner. They point at different next steps.

Record a voice message

1. Click Record a voice message.
2. Click Record, speak, then click Stop.
3. Click Save. The recorder shows Converting…, then Uploading… with a percentage, then Checking the recording…. The clip lands in the library, the recorder closes and the list refreshes.

Voice clips are capped at 2 minutes. If a save fails, the take stays in the recorder so you can click Save again. To stop without saving, click Close recorder.

Upload a file

Two places take a file:

• On the library page: in Add media, click Upload video, Upload audio or Upload document.
• In the flow builder: add a Voice message, Video or Document (PDF) block to a Message step, and click Upload in its picker.

Then:

1. Click Choose a file and pick it.
2. Wait for it: Converting… (only when the file needs converting), Uploading… with a percentage, then Checking the file….
3. It's added to the library. In the flow builder it's also selected for the block.

Closing the upload while a file is on its way cancels it, and nothing is saved. An upload that doesn't finish (a closed tab, a dropped connection) is never saved either — upload it again.

Accepted kinds:

HyperDM converts what it needs to: voice becomes AAC-in-MP4, video becomes H.264+AAC MP4, PDFs pass through. That's because Instagram and WhatsApp accept different formats, and the encode targets are the two shapes both channels take.

The caps are checked twice: in your browser before anything is sent, and again on the saved file. A file that isn't what its type says (a renamed file, say) is refused with "That file's contents don't match its type. Re-export it and try again."

Instagram accepts PDF and nothing else as a document. That is why there is no generic "file" upload — offering one would invite uploads the platform rejects.

Browse the library

Three tabs: Voice, Video, Documents.

Search with the box at the top — it searches names and transcripts, so you can find a voice clip by something said in it.

Each asset shows its status. A file that is still uploading shows Processing…. An asset that couldn't be saved says why in a few words (for example Not a valid file of its type), or Couldn't be converted. Only ready assets can be sent — a flow referencing an unfinished one is flagged as a go-live blocker rather than silently sending nothing.

Transcripts

Voice and video assets are transcribed automatically, so you can read a clip instead of playing it, and search across everything you've recorded.

If a clip can't be transcribed, a short note under the recorder or the upload says so — for example "We couldn't hear any speech in this one, so there's no transcript to search by. It'll still send fine." The clip is saved and sends either way.

Rename and delete

• Rename — click the name, type, press Enter (or Escape to cancel).
• Delete — the confirmation tells you whether the asset is used in a flow. If it isn't: "It isn't used in any flow. This can't be undone." If it is, the dialog lists the flows and warns "Those steps will stop sending it" — you can still delete, and a live flow that referenced it skips that step rather than breaking.

Storage

A meter at the top shows X of Y · N of M files. The bar turns red at 90%.

If an upload would take you over, the message says exactly which limit you hit and by how much — before you spend two minutes recording something that won't fit.

Use media in a flow

1. Open the flow builder.
2. Add a Message node.
3. Add a Voice message, Video or Document (PDF) block.
4. Pick the asset from your library.

TikTok DMs can't carry voice, video or documents — only text and quick replies. HyperDM can't send images on TikTok either: TikTok needs the picture uploaded to it first, and that step isn't built yet. A TikTok flow containing a media or image block is flagged while you edit.

Common questions

Why is the recorder button missing or disabled? Three possibilities: your plan doesn't include media (upgrade); your role can't upload (ask an owner); or your browser can't record — recording needs Chrome, Edge or a recent Safari, and the button says so when the browser is the problem. The page tells you which.

It says HyperDM needs microphone access. Recording uses your browser's microphone permission, and the browser asks the first time. If you said no, allow the microphone for this site in your browser's site settings and try again — or upload a file instead.

Do files I send in the Inbox show up here? No. A photo, video, voice message or PDF attached to an Inbox reply belongs to that conversation: it isn't listed here and doesn't count against your storage. Attachments you never send are deleted after 24 hours, and sent ones 30 days after sending. See The Inbox.

My upload says "Couldn't be converted". The source file wasn't in an accepted format, or the conversion failed. Re-export it as MP4 (video), M4A or WAV (voice), or PDF (document).

My upload stopped partway. Nothing was saved, so there's nothing to clean up — upload the file again. If it keeps stopping, check your connection: the file goes straight from your browser to storage, and the progress shows how far it got.

Can I send an image? Yes, but images aren't stored here — an Image block in a flow takes a pasted URL. In the Inbox you can attach a photo to a reply directly.

Can I search inside a voice message? Yes — the search box covers transcripts as well as names.