This guide describes how to use Copilot Chat to ask questions about software development in your IDE. You can ask general questions about software development, or specific questions about the code in your project. For more information, see About GitHub Copilot Chat.
You can ask Copilot Chat to give you code suggestions, explain code, generate unit tests, and suggest code fixes.
To open the chat view, click the icon in the title bar of Visual Studio Code. If the icon is not displayed, right-click the title bar and make sure that Command Center is selected.
Evaluate Copilot's response, and make a follow-up request if needed.
The response may contain text, code blocks, buttons, images, URIs, and file trees. The response often includes interactive elements. For example, the response may include a menu to insert a code block, or a button to invoke a Visual Studio Code command.
To see the files that Copilot Chat used to generate the response, select the Used n references dropdown at the top of the response. The references may include a link to a custom instructions file for your repository. This file contains additional information that is automatically added to all of your chat questions to improve the quality of the responses. For more information, see Adding repository custom instructions for GitHub Copilot.
Chat participants are like domain experts who have a specialty that they can help you with.
Copilot Chat can infer relevant chat participants based on your natural language prompt, improving discovery of advanced capabilities without you having to explicitly specify the participant you want to use in your prompt.
Note
Automatic inference for chat participants is currently in public preview and is subject to change.
Alternatively, you can manually specify a chat participant to scope your prompt to a specific domain. To do this, type @ in the chat prompt box, followed by a chat participant name.
Use slash commands to avoid writing complex prompts for common scenarios. To use a slash command, type / in the chat prompt box, followed by a command.
Copilot's GitHub-specific skills expand the type of information Copilot can provide. To access these skills in Copilot Chat, include @github in your question.
When you add @github to a question, Copilot dynamically selects an appropriate skill, based on the content of your question. You can also explicitly ask Copilot Chat to use a particular skill. You can do this in two ways:
Use natural language to ask Copilot Chat to use a skill. For example, @github Search the web to find the latest GPT model from OpenAI.
To specifically invoke a web search you can include the #web variable in your question. For example, @github #web What is the latest LTS of Node.js?
You can generate a list of currently available skills by asking Copilot: @github What skills are available?
You can change the large language model that Copilot uses to generate responses to chat prompts, including premium models with advanced capabilities. You may find that different models perform better, or provide more useful responses, depending on the type of questions you ask. For more information see Changing the AI model for GitHub Copilot Chat.
In addition to submitting prompts through the chat view, you can submit prompts in other ways:
Quick chat: To open the quick chat dropdown, enter ⇧+⌥+⌘+L (Mac) / Ctrl+Shift+Alt+L (Windows/Linux).
Inline: To start an inline chat directly in the editor or integrated terminal, enter Command+i (Mac) / Ctrl+i (Windows/Linux).
Smart actions: To submit prompts via the context menu, right click in your editor, select Copilot in the menu that appears, then select one of the actions. Smart actions can also be accessed via the sparkle icon that sometimes appears when you select a line of code.
Edit mode is only available in Visual Studio Code and JetBrains IDEs.
Use edit mode when you want more granular control over the edits that Copilot proposes. In edit mode, you choose which files Copilot can make changes to, provide context to Copilot with each iteration, and decide whether or not to accept the suggested edits after each turn.
Edit mode is best suited to use cases where:
You want to make a quick, specific update to a defined set of files.
You want full control over the number of LLM requests Copilot uses.
To start an edit session, select Open Chat from the Copilot Chat menu.
At the bottom of the chat panel, select Edit from the mode dropdown.
Optionally, add relevant files to the working set to indicate to GitHub Copilot which files you want to work on.
Submit a prompt. In response to your prompt, Copilot Edits determines which files in your working set to change and adds a short description of the change.
Review the changes and Apply or Discard the edits for each file.
For more detailed instructions, see Copilot Edits in the Visual Studio Code documentation.
Use agent mode when you have a specific task in mind and want to enable Copilot to autonomously edit your code. In agent mode, Copilot determines which files to make changes to, offers code changes and terminal commands to complete the task, and iterates to remediate issues until the original task is complete.
Agent mode is best suited to use cases where:
Your task is complex, and involves multiple steps, iterations, and error handling.
You want Copilot to determine the necessary steps to take to complete the task.
The task requires Copilot to integrate with external applications, such as an MCP server.
To start an edit session, select Open Chat from the Copilot Chat menu.
At the bottom of the chat panel, select Agent from the mode dropdown.
Submit a prompt. In response to your prompt, Copilot streams the edits in the editor, updates the working set, and if necessary, suggests terminal commands to run.
Review the changes. If Copilot suggested terminal commands, confirm whether or not Copilot can run them. In response, Copilot iterates and performs additional actions to complete the task in your original prompt.
For more information, see Copilot Edits in the Visual Studio Code documentation.
When you use Copilot agent mode, each prompt you enter counts as one premium request, multiplied by the model’s multiplier. For example, if you're using the included model—which has a multiplier of 0—your prompts won’t consume any premium requests. Copilot may take several follow-up actions to complete your task, but these follow-up actions do not count toward your premium request usage. Only the prompts you enter are billed—tool calls or background steps taken by the agent are not charged.
The total number of premium requests you use depends on how many prompts you enter and which model you select. See Requests in GitHub Copilot.
If you see the AI model picker at the bottom right of the chat view, select one of the models that supports adding images to prompts:
GPT-4.1 (the default that's used if you don't see a model picker)
Claude Sonnet 3.5
Claude Sonnet 3.7
Gemini 2.0 Flash
Gemini 2.5 Pro
Do one of the following:
Copy an image and paste it into the chat view.
Drag and drop one or more image file from your operating system's file explorer—or from the Explorer in VS Code—into the chat view.
Right-click an image file in the VS Code Explorer and click Copilot then Add File to Chat.
Type your prompt into the chat view to accompany the image. For example, explain this diagram, describe each of these images in detail, what does this error message mean.
Evaluate Copilot's response, and submit a follow up prompt if needed.
The response often includes interactive elements. For example, the response may include buttons to copy, insert, or preview the result of a code block.
To see the files that Copilot Chat used to generate the response, click the References link below the response. The references may include a link to a custom instructions file for your repository. This file contains additional information that is automatically added to all of your chat questions to improve the quality of the responses. For more information, see Adding repository custom instructions for GitHub Copilot.
GitHub Copilot Extensions integrate the power of external tools into Copilot Chat, helping you reduce context switching and receive responses with domain-specific context. You can install Copilot Extensions from the GitHub Marketplace or build private ones within your organization, then type @ in a chat window to see a list of your available extensions. To use an extension, select the extension from the list or type the full slug name, then type your prompt.
Use slash commands to avoid writing complex prompts for common scenarios. To use a slash command, type / in the chat prompt box, followed by a command.
By default, Copilot Chat will reference the file that you have open or the code that you have selected. You can also use # followed by a file name, file name and line numbers, or solution to reference a specific file, lines, or solution.
The @github chat participant is currently in preview, and only available in Visual Studio 2022 Preview 2 onwards.
Copilot's GitHub-specific skills expand the type of information Copilot can provide. To access these skills in Copilot Chat in Visual Studio, include @github in your question.
When you add @github to a question, Copilot dynamically selects an appropriate skill, based on the content of your question. You can also explicitly ask Copilot Chat to use a particular skill. For example, @github Search the web to find the latest GPT4 model from OpenAI.
You can generate a list of currently available skills by asking Copilot: @github What skills are available?
You can change the large language model that Copilot uses to generate responses to chat prompts, including premium models with advanced capabilities. You may find that different models perform better, or provide more useful responses, depending on the type of questions you ask. For more information see Changing the AI model for GitHub Copilot Chat.
In addition to submitting prompts through the chat window, you can submit prompts inline. To start an inline chat, right click in your editor window and select Ask Copilot.
This feature is currently in public preview and subject to change.
Available in Visual Studio 17.14 and later.
Copilot Edits lets you make changes across multiple files from a single Copilot Chat prompt
Use agent mode when you have a specific task in mind and want to enable Copilot to autonomously edit your code. In agent mode, Copilot determines which files to make changes to, offers code changes and terminal commands to complete the task, and iterates to remediate issues until the original task is complete.
In the Visual Studio menu bar, click View, then click GitHub Copilot Chat.
At the bottom of the chat panel, select Agent from the mode dropdown.
Submit a prompt. In response to your prompt, Copilot streams the edits in the editor, updates the working set, and if necessary, suggests terminal commands to run.
Review the changes. If Copilot suggested terminal commands, confirm whether or not Copilot can run them. In response, Copilot iterates and performs additional actions to complete the task in your original prompt.
When you use Copilot agent mode, each prompt you enter counts as one premium request, multiplied by the model’s multiplier. For example, if you're using the included model—which has a multiplier of 0—your prompts won’t consume any premium requests. Copilot may take several follow-up actions to complete your task, but these follow-up actions do not count toward your premium request usage. Only the prompts you enter are billed—tool calls or background steps taken by the agent are not charged.
If you see the AI model picker at the bottom right of the chat view, select one of the models that supports adding images to prompts:
GPT-4.1 (the default that's used if you don't see a model picker)
Claude Sonnet 3.5
Claude Sonnet 3.7
Gemini 2.0 Flash
Gemini 2.5 Pro
Do one of the following:
Copy an image and paste it into the chat view.
Click the paperclip icon at the bottom right of the chat view, click Upload Image, browse to the image file you want to attach, select it and click Open.
You can add multiple images if required.
Type your prompt into the chat view to accompany the image. For example, explain this image, or describe each of these images in detail.
To share feedback about Copilot Chat, you can use the Send feedback button in Visual Studio. For more information on providing feedback for Visual Studio, see the Visual Studio Feedback documentation.
In the top right corner of the Visual Studio window, click the Send feedback button.
Choose the option that best describes your feedback.
Evaluate Copilot's response, and submit a follow up prompt if needed.
The response often includes interactive elements. For example, the response may include buttons to copy or insert a code block.
To see the files that Copilot Chat used to generate the response, click the References link below the response. The references may include a link to a custom instructions file for your repository. This file contains additional information that is automatically added to all of your chat questions to improve the quality of the responses. For more information, see Adding repository custom instructions for GitHub Copilot.
Chat participants are like domain experts who have a specialty that they can help you with. You can use a chat participant to scope your prompt to a specific domain. To do this, type @ in the chat prompt box, followed by a chat participant name.
GitHub Copilot Extensions integrate the power of external tools into Copilot Chat, helping you reduce context switching and receive responses with domain-specific context. You can install Copilot Extensions from the GitHub Marketplace or build private ones within your organization, then type @ in a chat window to see a list of your available extensions. To use an extension, select the extension from the list or type the full slug name, then type your prompt.
Use slash commands to avoid writing complex prompts for common scenarios. To use a slash command, type / in the chat prompt box, followed by a command.
By default, Copilot Chat will reference the file that you have open or the code that you have selected. You can also tell Copilot Chat which files to reference by dragging a file into the chat prompt box. Alternatively, you can right click on a file, select GitHub Copilot, then select Reference File in Chat.
Copilot's GitHub-specific skills expand the type of information Copilot can provide. To access these skills in Copilot Chat, include @github in your question.
When you add @github to a question, Copilot dynamically selects an appropriate skill, based on the content of your question. You can also explicitly ask Copilot Chat to use a particular skill. You can do this in two ways:
Use natural language to ask Copilot Chat to use a skill. For example, @github Search the web to find the latest GPT model from OpenAI.
To specifically invoke a web search you can include the #web variable in your question. For example, @github #web What is the latest LTS of Node.js?
You can generate a list of currently available skills by asking Copilot: @github What skills are available?
You can change the large language model that Copilot uses to generate responses to chat prompts, including premium models with advanced capabilities. You may find that different models perform better, or provide more useful responses, depending on the type of questions you ask. For more information see Changing the AI model for GitHub Copilot Chat.
Built-in requests. In addition to submitting prompts through the chat window, you can submit built-in requests by right clicking in a file, selecting GitHub Copilot, then selecting one of the options.
Inline. You can submit a chat prompt inline, and scope it to a highlighted code block or your current file.
To start an inline chat, right click on a code block or anywhere in your current file, hover over GitHub Copilot, then select Copilot: Inline Chat, or enter Ctrl+Shift+I.
Edit mode is only available in Visual Studio Code and JetBrains IDEs.
Use edit mode when you want more granular control over the edits that Copilot proposes. In edit mode, you choose which files Copilot can make changes to, provide context to Copilot with each iteration, and decide whether or not to accept the suggested edits after each turn.
Edit mode is best suited to use cases where:
You want to make a quick, specific update to a defined set of files.
You want full control over the number of LLM requests Copilot uses.
To start an edit session, click Copilot in the menu bar, then select Open GitHub Copilot Chat.
At the top of the chat panel, click Copilot Edits.
Add relevant files to the working set to indicate to GitHub Copilot which files you want to work on. You can add all open files by clicking Add all open files or individually search for single files.
Submit a prompt. In response to your prompt, Copilot Edits determines which files in your working set to change and adds a short description of the change.
Review the changes and Accept or Discard the edits for each file.
Use agent mode when you have a specific task in mind and want to enable Copilot to autonomously edit your code. In agent mode, Copilot determines which files to make changes to, offers code changes and terminal commands to complete the task, and iterates to remediate issues until the original task is complete.
Agent mode is best suited to use cases where:
Your task is complex, and involves multiple steps, iterations, and error handling.
You want Copilot to determine the necessary steps to take to complete the task.
The task requires Copilot to integrate with external applications, such as an MCP server.
To start an edit session using agent mode, click Copilot in the menu bar, then select Open GitHub Copilot Chat.
At the top of the chat panel, click the Agent tab.
Submit a prompt. In response to your prompt, Copilot streams the edits in the editor, updates the working set, and if necessary, suggests terminal commands to run.
Review the changes. If Copilot suggested terminal commands, confirm whether or not Copilot can run them. In response, Copilot iterates and performs additional actions to complete the task in your original prompt.
When you use Copilot agent mode, each prompt you enter counts as one premium request, multiplied by the model’s multiplier. For example, if you're using the included model—which has a multiplier of 0—your prompts won’t consume any premium requests. Copilot may take several follow-up actions to complete your task, but these follow-up actions do not count toward your premium request usage. Only the prompts you enter are billed—tool calls or background steps taken by the agent are not charged.
The total number of premium requests you use depends on how many prompts you enter and which model you select. See Requests in GitHub Copilot.
Evaluate Copilot's response, and submit a follow up prompt if needed.
The response often includes interactive elements. For example, the response may include buttons to copy or insert a code block.
To see the files that Copilot Chat used to generate the response, click the References link below the response. The references may include a link to a custom instructions file for your repository. This file contains additional information that is automatically added to all of your chat questions to improve the quality of the responses. For more information, see Adding repository custom instructions for GitHub Copilot.
You can change the large language model that Copilot uses to generate responses to chat prompts, including premium models with advanced capabilities. You may find that different models perform better, or provide more useful responses, depending on the type of questions you ask. For more information see Changing the AI model for GitHub Copilot Chat.
Use slash commands to avoid writing complex prompts for common scenarios. To use a slash command, type / in the chat prompt box, followed by a command.
To see all available slash commands, type / in the chat prompt box. For more information, see GitHub Copilot Chat cheat sheet.
Use agent mode when you have a specific task in mind and want to enable Copilot to autonomously edit your code. In agent mode, Copilot determines which files to make changes to, offers code changes and terminal commands to complete the task, and iterates to remediate issues until the original task is complete.
Agent mode is best suited to use cases where:
Your task is complex, and involves multiple steps, iterations, and error handling.
You want Copilot to determine the necessary steps to take to complete the task.
The task requires Copilot to integrate with external applications, such as an MCP server.
To open the chat view, click Copilot in the menu bar, then click Open Chat.
At the bottom of the chat panel, select Agent from the mode dropdown.
Optionally, add relevant files to the working set view to indicate to Copilot which files you want to work on.
Submit a prompt. In response to your prompt, Copilot streams the edits in the editor, updates the working set, and if necessary, suggests terminal commands to run.
Review the changes. If Copilot suggested terminal commands, confirm whether or not Copilot can run them. In response, Copilot iterates and performs additional actions to complete the task in your original prompt.
When you use Copilot agent mode, each prompt you enter counts as one premium request, multiplied by the model’s multiplier. For example, if you're using the included model—which has a multiplier of 0—your prompts won’t consume any premium requests. Copilot may take several follow-up actions to complete your task, but these follow-up actions do not count toward your premium request usage. Only the prompts you enter are billed—tool calls or background steps taken by the agent are not charged.
The total number of premium requests you use depends on how many prompts you enter and which model you select. See Requests in GitHub Copilot.
By default, Copilot Chat will reference the file that you have open or the code that you have selected. To attach a specific file as reference, click in the chat prompt box.
You can open a conversation thread for each Xcode IDE to keep discussions organized across different contexts. You can also revisit previous conversations and reference past suggestions through the chat history.
Compatible version of Eclipse. To use the GitHub Copilot extension, you must have Eclipse version 2024-09 or above. See the Eclipse download page.
If you are a member of an organization or enterprise with a Copilot Business or Copilot Enterprise plan, the "MCP servers in Copilot" policy must be enabled in order to use MCP with Copilot.
Use slash commands to avoid writing complex prompts for common scenarios. To use a slash command, type / in the chat prompt box, followed by a command. For example, use /explain to ask Copilot to explain the code in the file currently displayed in the editor.
To see all available slash commands, type / in the chat prompt box.
You can change the large language model that Copilot uses to generate responses to chat prompts, including premium models with advanced capabilities. You may find that different models perform better, or provide more useful responses, depending on the type of questions you ask. For more information see Changing the AI model for GitHub Copilot Chat.
Use agent mode when you have a specific task in mind and want to enable Copilot to autonomously edit your code. In agent mode, Copilot determines which files to make changes to, offers code changes and terminal commands to complete the task, and iterates to remediate issues until the original task is complete.
Agent mode is best suited to use cases where:
Your task is complex, and involves multiple steps, iterations, and error handling.
You want Copilot to determine the necessary steps to take to complete the task.
The task requires Copilot to integrate with external applications, such as an MCP server.
To open the Copilot Chat panel, click the Copilot icon () in the status bar at the bottom of Eclipse, then click Open Chat.
At the bottom of the chat panel, select Agent from the mode dropdown.
Submit a prompt. In response to your prompt, Copilot streams the edits in the editor, updates the working set, and if necessary, suggests terminal commands to run.
Review the changes. If Copilot suggested terminal commands, confirm whether or not Copilot can run them. In response, Copilot iterates and performs additional actions to complete the task in your original prompt.
When you use Copilot agent mode, each prompt you enter counts as one premium request, multiplied by the model’s multiplier. For example, if you're using the included model—which has a multiplier of 0—your prompts won’t consume any premium requests. Copilot may take several follow-up actions to complete your task, but these follow-up actions do not count toward your premium request usage. Only the prompts you enter are billed—tool calls or background steps taken by the agent are not charged.
The total number of premium requests you use depends on how many prompts you enter and which model you select. See Requests in GitHub Copilot.
