Scrapeer
Integrations

AI Block

AI sends prompt input to a configured AI provider and saves the parsed result into a variable.

At a glance

Best for
Language tasks
Modes
Summarize, classify, extract, score, custom
Requires
AI connection
Error handle
Yes

What it does

AI is for language-level processing after Scrapeer has collected page or variable content.

It can return text, objects, scores, or categories depending on mode and schema.

Use it when

  • Summarize extracted page content.
  • Classify listings or messages into categories.
  • Extract structured fields from unstructured text.

Do not use it when

  • A selector or deterministic parser can extract the value directly.
  • The data is sensitive and should not be sent to the configured provider.
  • The Flow needs guaranteed deterministic output without model variance.

Settings

AI needs a connection, mode, source text or custom prompt, model settings, and output variable.

Connection

Required: Yes
Variables
No
Description
Configured AI provider connection.

Mode

Required: Yes
Variables
No
Description
Summarize, classify, extract, score, or custom.

Source variable

Required: Except custom mode
Variables
Variable name
Description
Input text or object to process.

Prompt and schema

Required: Depends on mode
Variables
Yes
Description
Instructions, labels, fields, score bounds, or custom schema.

Output variable

Required: Yes
Variables
Variable name
Description
Stores the AI result.

Outputs

AI output depends on mode.

{{outputVariableName}}string | object | number

Parsed AI result.

categorystring

Classification label when using classify mode.

Example Flow

Use AI after extraction when the next step needs interpretation rather than raw text.

  1. Extract page content or rows.
  2. Run AI with the appropriate mode.
  3. Use the result in Conditional, Template, or export Blocks.

Errors and fixes

Start with symptoms visible in the editor or Inspector, then check the earliest Block that produced the wrong page state, variable, or output.

The AI result is empty or malformed.

Check
Check source variable and prompt/schema settings.
Fix
Use a smaller, clearer source value and a stricter output schema.

The Block fails before calling the model.

Check
Check connection and required input fields.
Fix
Configure a valid connection and source variable.

Local and cloud runs

AI behavior depends on the configured provider connection and model. Keep prompts and schemas small enough for reliable output.

Blocks

Understand how Blocks connect and pass state through a Flow.

Runs and outputs

Inspect Block output, errors, and variables after a run.