Motionlab

Motionlab renders one personalized video per recipient, composing their name, photo, location, or offer into the footage. Agents generate videos from a campaign, poll render status, read campaign parameters, and run test mode before spending credits on a real send.

Verified 6 actions API key available

Motionlab website Platform Documentation Capability data verified 2026-08-28

A list of accounts is approved for a personalized video touch

Get Campaign Parameters returns every personalization field with its type and whether it is required

Agent maps each CRM field to a campaign key and confirms that every image value is a publicly reachable JPG or PNG

Generate Video runs the whole list in test mode, which produces the same links and spends nothing

Get Render Status confirms each test came back as a test render rather than a failure

The campaign owner gets the sampled player links beside the field mapping that produced them

The campaign owner authorizes the real run, and only then does the agent send the list without the test flag

What This Integration Enables

Motionlab renders one video per recipient, composing their name, their photo, their city, or their offer into the footage itself. That makes it a different kind of tool from a video generator, because the unit of work is a person rather than a video, and the failure mode is different too: a mistake here is not an ugly asset, it is a named individual receiving a video that gets their name, their company, or their location wrong.

The connector is deliberately narrow, and the shape follows from one fact: the API key is the campaign. Motionlab issues one key per campaign at its launch step, which is why the campaign operations take no campaign identifier at all and why one connection addresses exactly one campaign. FlowRunner agents read the campaign's personalization fields with their types and requirements, generate a video per recipient, poll for render status, and check whether the campaign's endpoint is actually launched, which is what separates a bad key from a campaign that was never turned on.

Without FlowRunner

Personalized video is a project Each recipient's video is produced by hand, so the list is short enough to be worth producing and too short to matter

Mapping errors reach recipients A field mapped to the wrong key composes the wrong name into footage that arrives addressed to someone by name

Spend is discovered afterward Credits are consumed by a run nobody previewed, and the first review is of a video already delivered

With FlowRunner

Personalized video is a list operation The unit of work is a recipient, and a list of two thousand costs the same effort as a list of ten

Mapping is checked against the campaign Field keys and types are read from the campaign itself before a render is attempted, not guessed from a spreadsheet header

Spend follows a review The list is rendered in test mode first, and a person releases the run that actually bills

Use Case Scenarios

The account list gets a video each

A target list is approved in HubSpot. The agent calls Get Campaign Parameters to learn the exact keys the campaign expects and their types, maps company name, first name, and the account's city onto them, and confirms that the logo URL held on each record is publicly reachable, because Motionlab returns an error rather than a placeholder when it cannot fetch an image. It renders the list in test mode, posts the sample for review, and after approval runs it for real. Each player link is written back to the contact record, and the sequence in Instantly picks it up as a merge field.

A campaign that has not actually been launched

A scheduled run starts failing with an authorization error. Rather than assuming a rotated key, the agent calls Get API Status, which is what distinguishes a wrong credential from a campaign whose endpoint was never launched. The answer goes into Slack with the campaign name from Get Campaign Info, so the person who owns the campaign is told which of the two problems they have instead of being handed a generic failure.

Following a render without a subscription to register

Motionlab posts to a callback URL configured inside the campaign rather than through the API, so there is no subscription for a workflow to create or clean up. For a flow that would rather not depend on a dashboard setting, the agent polls Get Render Status per video, reads the failure reason when one comes back, and retries only the recipients whose renders actually failed rather than the whole list.

Human-in-Loop Highlight

Motionlab built the pause into its own API, which is the strongest signal there is that this is where a person belongs. Test mode returns the same response shape, the same player and thumbnail links, and fires the same callback as a real run, and it renders nothing and costs nothing. A real run bills per render and produces a video with a specific person's name composed into the footage, which cannot be recalled once the link is in an email. So the agent renders the list in test mode, pulls a handful of samples, and stops: "Campaign Q3 Outbound, 1,840 recipients. Field mapping is company to accountName, city to place, and logo to imageUrl. Here are five sample renders and the records behind them. 12 records have no logo URL and will render with the campaign default. Release the real run?" The campaign owner checks five videos and answers. Everything before that point is machine work, and the one decision that spends credits and puts a stranger's name in front of them stays with a person. That is what FlowRunner's human-in-the-loop step is for.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via email

Human decides

Agent resumes with decision

Agent Capabilities

6 actions

Rendering

Generate Video Renders one personalized video for one recipient from the campaign's personalization fields, returning the video id and its player and thumbnail links before rendering has finished. Its test mode flag produces the same response and the same links while rendering nothing and spending nothing, which is what makes a full dry run of a list possible.
Get Render Status Returns whether a render is running, finished, failed, or was a test, along with the failure reason when there is one. The operation a flow polls, and the one that lets a retry target only the recipients that actually failed.
Get Video Info Returns a rendered video's state and its duration in seconds.

Campaign

Get Campaign Info Returns the campaign this connection addresses, including its name, id, thumbnail, preset, and status. It takes no campaign identifier, because the key already names one.
Get Campaign Parameters Returns every personalization field the campaign accepts, with its key, label, type, whether it is required, and its options where it is an enumeration. The read that belongs before any generation call, because guessing keys produces an error rather than a default.
Get API Status Reports whether the campaign's endpoint has been launched. This is what separates a wrong key from a campaign that was never turned on, since Motionlab answers both the same way.

Frequently Asked Questions

What can FlowRunner do with Motionlab?

FlowRunner agents can run Generate Video, Get Render Status, and Get Video Info in Motionlab, plus 3 more actions.

Does connecting Motionlab to FlowRunner require OAuth?

No. Motionlab connects to FlowRunner with an API key, no OAuth flow required.

Can Motionlab trigger a FlowRunner workflow automatically?

Motionlab doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.

Start building with Motionlab

$100 in credits. No card required. Connect in minutes.

Start Building Free Talk to us