How to retrieve the result object for a specific Phantom launch via API

To retrieve the result object for a specific Phantom launch, call GET /containers/fetch-result-object with the Container ID of the run you want. The API returns the output data directly in the response as a resultObject field, not as a downloadable file. Use this when you need data from one specific run rather than accumulated result files.

Before you start

Before retrieving the result object for a specific launch, make sure:

• The Phantom has been launched at least once.
• You have a valid API key.
  → If you don't have one, see Create an API key. API keys can only be created by Workspace admins, if you don't see the Add API key button, check with your admin.
• You know the Launch ID - also called Container ID - for the run you want to retrieve results from.

→ If any of these are missing, start with How to Get Started with the PhantomBuster API and How to find your Container ID for API calls.

How launch-specific result retrieval works

Each time a Phantom runs, PhantomBuster creates a new launch (also called a container).

• That launch has its own Container ID.
• Each launch produces its own output.

For most Phantoms, the API returns the result data directly in the response and does not generate CSV or JSON files for a specific launch. In some edge cases, when the output is very large, the API may return file references instead of inline data.

Instead of working with accumulated files, this endpoint lets you retrieve the result object associated with one specific launch.

The result object contains the actual output data produced by the Phantom during that run, returned directly in the API response.

→ This is the method to use when you need launch-level data, not accumulated files.

Retrieve the result object from the API reference

Step 1: Open the launch result endpoint

1. Open the PhantomBuster API reference.
2. In the left menu, go to:
   PhantomBuster API v2 → Containers → GET /containers/fetch-result-object

Step 2: Add your API key

1. In the top-right corner of the page, locate the Credentials section.
2. Paste your API key into the X-Phantombuster-Key header.

Step 3: Add the Container ID

1. In the Query Params section, locate the field named id.
2. Paste the Container ID of the Phantom run you want results for (see How to Find your Container ID for API Calls).

Step 4: Send the request

• Click Try It!

Understand the API response

If the request is successful:

• The API returns a 200 status.
• The response includes a resultObject field.

Depending on the Phantom, it may include:

• Rows of extracted data
• Records or objects generated during the run
• This data corresponds only to the launch you requested, not to other runs.

This corresponds to what you can see in the PhantomBuster dashboard when inspecting a launch’s result data, but without creating CSV or JSON files.

Important limitations

• This endpoint returns data, not downloadable files.
• For most Phantoms, the response contains the data directly (via the resultObject field).
• In some edge cases, when the output is very large, the response may include file references instead of inline data.
• If a launch did not generate results, the response may be empty even if the request succeeds.

→ If your goal is to download CSV or JSON result files, use How to retrieve a Phantom’s result files (CSV or JSON) via API instead.

Frequently asked questions

What's the difference between this endpoint and /agents/fetch?

The /agents/fetch endpoint gives you file locations for the Phantom's accumulated result files (CSV/JSON). The /containers/fetch-result-object endpoint returns the actual output data from one specific launch, inline in the response.

Does this endpoint generate a CSV or JSON file?

No. It returns the data directly in the API response. If you need downloadable files, use the result files endpoint (/agents/fetch) instead.

What if the response is empty?

If the launch didn't generate any results (for example, if it failed or found no data), the response may be empty even with a 200 status.

Can the response contain file references instead of data?

In rare cases with very large outputs, yes. The response may include file references instead of inline data.