AI agents, automations and apps for VLM Run
Set up an automation: when new message in channels in Slack, image to JSON in VLM Run
Build itVLM Run works as a step: another app’s event or a schedule starts it.
What you can build on VLM Run
Slack to VLM Run, every time
When new message in channels in Slack, image to JSON in VLM Run, the same way every time.
An AI agent that works in VLM Run
It uses “List models or domains”, “Get predictions” and “List files” in VLM Run when your instructions call for it.
Pages on vlm.run
For VLM Run’s public pages on vlm.run that its triggers and actions don’t reach: it reads them on a schedule and reports what changed.
Your own jsons app
Your VLM Run jsons as records, with a board and a table, kept current with “Image to JSON”.
VLM Run triggers and actions
A trigger starts the work; actions are the steps in VLM Run.
Actions
18- Image to JSONGenerate structured prediction for the given image.
- Convert document to JSONGenerate structured JSON from a document; configure domain, custom JSON schema, or agent skills.
- List models or domainsRetrieve available models or domains by selecting view mode (models or domains).
- Get predictionsRetrieve one or multiple predictions, optionally by prediction ID or using pagination.
- List filesRetrieve a file by ID or list files; configure skip and limit pagination.
- Convert audio to JSONConvert audio files or URLs into structured JSON; configure skills, custom schema, and callback.
- Generate video JSONGenerate JSON from a video URL or uploaded file. Configure model, video settings, skills, advanced options and metadata.
- Create agentCreate a new agent with a prompt, model, tools, callbacks, and optional skills or advanced settings.
- Upload fileUpload a file to VLM Run and return the file metadata including an optional public URL.
- Execute agentRun an agent to create a prediction, configuring model, inputs, prompt, skills, callback, and metadata.
- List agentsList agents with an optional result limit and selectable response fields.
- Lookup agentLookup an agent by ID or name; provide agent ID or name and optionally include a prompt.
- Create skillCreate a new Skill by uploading a file, using a prompt, or a chat session; set name, description, and visibility.
- List skillsRetrieve skills with pagination, sorting, and grouping options.
- Generate chat completionGenerate a chat completion from messages; configure model, sampling, tools, and response format.
- List agent executionsRetrieve an agent execution by ID or list executions with optional pagination and selectable fields.
- Update skillUpdate a skill's file, description, or public visibility by selecting the skill and inputs.
- Get skill download URLGet a presigned download link for a skill's zip file; pick the skill and optionally set how long the link lasts.
No actions match.
Templates for VLM Run
Pick one and Captain builds it with you.


When new message in channels in Slack, image to JSON in VLM Run
Build this

When new opportunity in Salesforce, image to JSON in VLM Run
Build this

When a new subscriber is added to an audience in MailChimp, image to JSON in VLM Run
Build this

When new form response in Google Forms, image to JSON in VLM Run
Build this

When new deal in Pipedrive, image to JSON in VLM Run
Build this

When new contact added to list in ActiveCampaign, image to JSON in VLM Run
Build this

When a form is submitted in Typeform, image to JSON in VLM Run
Build this

When new submission in Jotform, image to JSON in VLM Run
Build this

When new customer created in Square, image to JSON in VLM Run
Build this
Apps that work best with VLM Run
Questions about VLM Run in 50agents
What can 50agents do in VLM Run?
Its actions: “Image to JSON”, “Convert document to JSON”, “List models or domains”, “Get predictions”, “List files”, “Convert audio to JSON”, “Generate video JSON” and “Create agent”, and the rest listed above. Anything that reaches people outside asks you first.
Which apps is VLM Run most often paired with?
Slack, Salesforce, MailChimp, Google Forms and Pipedrive, in viaSocket’s templates.
What is VLM Run?
VLM Run is a platform designed to enhance your running experience by providing tools and resources for tracking, analyzing, and improving your running performance.
What do people build most with VLM Run?
The most used templates: when new message in channels in Slack, image to JSON in VLM Run; when new opportunity in Salesforce, image to JSON in VLM Run; when a new subscriber is added to an audience in MailChimp, image to JSON in VLM Run.
Can an AI agent work in VLM Run?
Yes. Connect VLM Run and it can use “List models or domains”, “Get predictions” and “List files” when its instructions call for it. It runs on a schedule or when you ask in its chat.
Does 50agents change my VLM Run data on its own?
Only through the actions you name, such as “Image to JSON”. Changes made in VLM Run cannot be undone from 50agents, and anything that reaches people outside your business asks you first.