Help Center/ CodeArts Agent/ User Guide (IDE)/ Context/ Multimodal Input
Updated on 2026-07-29 GMT+08:00
View PDF
Share

Multimodal Input

CodeArts Agent supports uploading images and other attachments to integrate them into the chat context. You can directly upload or drag and drop error screenshots, code blocks, or technical documentation into the input box. This eliminates the need for lengthy text descriptions, allowing you to clearly reproduce the problem scenario and achieve precise, efficient communication of your requirements.

Constraints

Table 1 Constraints

Category

Description

Agent mode
  • Uploading images from a local directory: You can upload 1 to 5 images at a time. Each image must not exceed 10 MB, and the total size cannot exceed 20 MB. Supported formats: PNG, JPG, JPEG, and WebP.
  • Uploading other attachments from a local directory: You can upload 1 to 20 items at a time with no file size limits. Supported formats: XLS, XLSX, CSV, DOC, DOCX, PPT, PPTX, PDF, MD, MDX, MDC, TXT, and SVG.
  • Dragging and dropping images or other files from the project directory

    Find the target image or file in the file tree under the explorer on CodeArts Agent IDE, and drag it into the input box to upload it.

Intelligent Q&A mode

Only local images can be uploaded.

  • Uploading images from a local directory: You can upload 1 to 5 images at a time. Each image must not exceed 4 MB, and the total size cannot exceed 4 MB. Supported formats: PNG, JPG, JPEG, and WebP.
  • Dragging and dropping images or other files from the project directory

    Find the target image or file in the file tree under the explorer on CodeArts Agent IDE, and drag it into the input box to upload it.

OS

The voice input function is supported only on the following OSs:

  • Windows OS: Windows 11 (x64) is recommended. For Windows 10 (x64), the version must be 2019 or later, and upgrading to the latest stable version is advised.
  • macOS: macOS 11 or later, compatible with ARM64 (Apple Silicon).

Uploading Images in a Chat

Example: uploading a code block screenshot to the agent.

  1. Switch to the agent mode.

    1. Log in to CodeArts Agent IDE as instructed in Quick Start.
    2. At the bottom of the input box on the chat panel, choose Built-in > Agent to switch to the agent mode. The selected model is displayed on the right. You can select a model as required from the drop-down list.

      If the chat panel of CodeArts Agent is not displayed, click (Expand AI Sidebar) in the upper right corner of the top menu bar to open it.

  2. Upload an image.

    • Images from a local directory

      Click in the input box, select Upload Attachments, and upload a local image. You can also drag and drop a local image into the input box to upload it.

    • Images from the clipboard

      Paste into the input box to upload an image from your clipboard.

    • Images from the project directory

      Find the target image in the file tree under the explorer on CodeArts Agent IDE, and drag it into the input box to upload it.

  3. View the image recognition results.

Uploading Other Attachments in a Chat

Example: uploading a PDF file to the agent.

  1. Switch to the agent mode.

    1. Log in to CodeArts Agent IDE as instructed in Quick Start.
    2. At the bottom of the input box on the chat panel, choose Built-in > Agent to switch to the agent mode. The selected model is displayed on the right. You can select a model as required from the drop-down list.

      If the chat panel of CodeArts Agent is not displayed, click (Expand AI Sidebar) in the upper right corner of the top menu bar to open it.

  2. Click in the input box, select Upload Attachments, and upload a local attachment.
  3. Enter a prompt and click . The agent will then create and execute a task based on the prompt.

Using Voice Input

Enable voice input when typing is inconvenient, when you need quick verbal communication, or when structuring text instructions is difficult. This function supports up to 10 minutes of continuous recording each time, eliminating lengthy text descriptions and helping you achieve precise and efficient communication of your needs.

When using this function for the first time, ensure that microphone access is enabled. If it is disabled, configure it in your system settings. (The exact interface may vary depending on your OS.)

  • Windows: Go to Start > Settings > Privacy & security and allow microphone access.
  • macOS: Go to System Settings > Privacy & Security and allow microphone access.
  1. Switch to the agent mode.

    1. Log in to CodeArts Agent IDE as instructed in Quick Start.
    2. At the bottom of the input box on the chat panel, choose Built-in > Agent to switch to the agent mode. The selected model is displayed on the right. You can select a model as required from the drop-down list.

      If the chat panel of CodeArts Agent is not displayed, click (Expand AI Sidebar) in the upper right corner of the top menu bar to open it.

  2. At the bottom of the input box on the chat panel, click and start recording. The AI will automatically convert your voice input into text and enter it in the input box.

    The duration of a single voice input cannot exceed 10 minutes.

  3. Once you finish voice input, click and then . The agent will automatically create a task based on your prompt.

    If there are errors in the voice recognition result, you can edit the text manually before sending.

Parent Topic: Context