Skip to content
← Back to documentation
Technical Reference

Anthropic (Claude)

Invoke Anthropic's Claude models to generate, summarize, or classify text and structured data within a workflow. Send a prompt and optional system message to get back a text or JSON response for use in downstream steps.

CategoryAI
AuthenticationAPI Key
Published scopes0 published scopes
Last updatedAugust 25, 2026
Use cases
Marketing Automation, Product Analytics, Revenue Operations, Support Ticket Sync
Industries
E-commerce & Retail, Media & Entertainment, Professional Services, SaaS

Overview

The Anthropic connector provides direct access to Claude language models within a flows360 workflow. This allows for the integration of generative AI capabilities, such as text generation, summarization, classification, and structured data extraction, into automated processes.

The primary action, Invoke Model, sends a prompt to a specified Claude model (e.g., Claude 3.5 Sonnet, Claude 3 Opus) and returns the model's response. This enables dynamic content creation, data analysis, and decision-making based on natural language inputs. The connector accepts a user prompt and an optional system message to set context for the model, influencing its tone and response style.

Responses can be returned as plain text for simple generation tasks or as structured JSON, which is essential for reliably passing data to downstream workflow steps. This makes it possible to extract specific entities from unstructured text, such as parsing contact information from an email or categorizing customer feedback into a predefined schema.

Capability summary

AuthenticationAPI Key
Supported objectsNot published
Supported actions3

Authentication

Authentication method: API Key

Supported actions

ActionDescription
Invoke ModelSends a prompt to a specified Claude model and returns the generated text or structured JSON output. Supports system messages for setting context.
Classify TextAnalyzes input text and assigns it to one or more predefined categories. Useful for sentiment analysis, topic modeling, or intent recognition.
Extract Structured DataParses unstructured text to extract specific entities or information into a structured JSON object, based on a user-defined schema.

Setup

  1. Log in to your Anthropic Console account at console.anthropic.com.
  2. Navigate to the API Keys section from the main menu.
  3. Click the Create Key button. Provide a descriptive name for the key (e.g., "flows360-integration") and click Create Key.
  4. Your API key will be displayed. Copy this key immediately, as it will not be shown again for security reasons.
  5. In your flows360 workspace, navigate to Connections and click Add Connection.
  6. Select Anthropic (Claude) from the list of available applications.
  7. In the configuration panel, give the connection a unique name.
  8. In the API Key field, paste the key you copied from the Anthropic Console.
  9. Click Save to create the connection. The connector is now ready to use in your workflows.

Limitations

  • Each Invoke Model action is a stateless, independent API call. The connector does not automatically manage conversation history. To maintain context across multiple calls, the previous prompts and responses must be explicitly passed in subsequent prompts.
  • The connector is subject to the rate limits and token quotas of your Anthropic API plan. Exceeding these limits will result in API errors. Refer to Anthropic's documentation for details on model-specific context window sizes and rate limits.
  • When requesting structured JSON output, the model's ability to generate valid, well-formed JSON that adheres to your requested schema is powerful but not guaranteed. It is recommended to include validation and error handling in workflows that consume this output.

Use cases and industries

Use cases

Marketing AutomationProduct AnalyticsRevenue OperationsSupport Ticket Sync

Industries

E-commerce & RetailMedia & EntertainmentProfessional ServicesSaaS