Skip to main content

Tutorial: Image Analysis

Goal: Extract text from an error dialog image that cannot be copied. Level: Intermediate Time: 1 minute

Prerequisites​

  • Vision-capable AI provider: Image analysis requires a provider that supports vision/multimodal capabilities:
    • OpenAI: GPT-5.4, GPT-5.5 (GPT-5.5 requires Pro Plus or higher)
    • Anthropic: Claude Opus 4.8 (requires Pro Plus or higher), Claude Sonnet 4.6, Claude Haiku 4.5
    • Google: Gemini 3.1 Pro, Gemini 3.5 Flash
    • xAI: Grok 4.3

Note: If your configured provider does not support image analysis, screenshot commands will not be available. Configure a vision-capable provider in Settings before using this feature.

Scenario​

An application crashed and showed a popup error message. You can't select the text with your mouse to Google it.

Step 1: Capture​

  1. Press the Screenshot Hotkey (macOS Shift+Cmd+S, Windows/Linux Ctrl+Shift+Alt+S).
  2. The screen freezes and crosshairs appear.
  3. Click and drag a box around the error dialog.

Step 2: Analyze​

Once you release the mouse, the Rephlo Screenshot Analysis Result dialog appears.

  1. Select the "Extract Text" command.
    • Instruction: "You are an OCR tool. Extract all visible text exactly as it appears."

Step 3: Result​

Rephlo returns the raw text:

"Error 0x8004005: Unspecified error in module lib_core.dll at address 0x4f..."

Step 4: Solve​

Now you can:

  1. Copy the error code.
  2. Run a "Troubleshoot" command on that extracted text to ask the AI how to fix it immediately.