GuidesRecipesAPI ReferenceChangelogDiscussions
Log In

Create a job

This endpoint submits an inference job request for any of the following data types:

  • a string of text
  • embedded data (via Base64 encoding) with a URI compatible string
  • database entries or files
  • files stored in AWS S3, Azure Blob Storage, or NetApp StorageGRID

Formatting the input Body Parameter

The input body parameter can be customized to allow for the processing of many different data sources against numerous model types. It must be formatted correctly for the type of input and the type of model that is being used.

Basic input body parameter format

"input": {
    "type": "input type",
    #additional json fields are type-dependent
    "sources": {
      "user-defined-key": { #This key can be any text
        "model-defined-key": { #This key is defined by the model developer and must match the model
          #Individual inputs are model-dependent and type-dependent

Valid type values

There are seven valid values for type field: text, embedded, jdbc, aws-s3, aws-s3-folder, azureblob, storage-grid.

Modifying the user-defined-key

The user-defined-key, as seen in the example input body parameter above, can be modified to be any utf-8 string. This key provides a convenient way for developers to provide unique names for each input being processed by Modzy. Examples might include: "image-1", "final_report.pdf", or "".

Modifying the model-defined-key

The model-defined-key, as seen in the example input body parameter above, is the key expected by the model that your input is being submitted to. This key is set by the model developer and must match what the model developer defined exactly. This value is found within each model's API page. Keys may look something like: "input.txt", "image", or "input.mp4".

Inference Job Examples

"type": "text": Plain Text Inputs
"type": "embedded": Embedded File Inputs
"type": "jdbc": Inputs from Databases
"type": "aws-s3": AWS S3 Inputs
"type": "aws-s3-folder": Entire AWS S3 Folder Inputs
"type": "azure-blobstore": Azure Blob Storage Inputs
"type": "storage-grid": NetApp StorageGRID Inputs


Auto-generated queries are not valid for this endpoint

The API response generator seen on the right sidebar cannot create valid queries for the /jobs endpoint because the body parameters require too much customization. Instead, click on the links above or view the HTTP examples provide for examples of value /jobs submissions.


Quota limits may block job requests. They define the maximum number of job inputs a user, team, or project can submit within a given date range.

Quota accumulation takes effect immediately when the endpoint submits a job. Modzy credits the input usage back (subtract from the accumulated value) when an input fails or the user submits a job cancelation.


Data Scientists, developers, and projects can submit jobs to visible purchased models with active versions.