[[["易于理解","easyToUnderstand","thumb-up"],["解决了我的问题","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["很难理解","hardToUnderstand","thumb-down"],["信息或示例代码不正确","incorrectInformationOrSampleCode","thumb-down"],["没有我需要的信息/示例","missingTheInformationSamplesINeed","thumb-down"],["翻译问题","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["最后更新时间 (UTC):2025-08-25。"],[],[],null,["# Create visualizations with Gemini assistance\n\n\u003cbr /\u003e\n\n|\n| This product or feature is subject to the \"Pre-GA Offerings Terms\" in the General Service Terms section\n| of the [Service Specific Terms](/terms/service-terms#1).\n|\n| Pre-GA products and features are available \"as is\" and might have limited support.\n|\n| For more information, see the\n| [launch stage descriptions](/products#product-launch-stages).\n\nThis documentation page describes how to use Gemini assistance to generate JSON formatting options from text-based prompts to accelerate the customization of Looker visualizations. Customizing visualizations with the **Visualization Assistant** is a Gemini in Looker feature that is available in Looker (original) and Looker (Google Cloud core) instances.\n\nThe **Visualization Assistant** is available for visualizations that use the HighCharts API, which includes most [Cartesian charts](/looker/docs/visualization-types#cartesian_charts), such as the [column chart](/looker/docs/column-options), [bar chart](/looker/docs/bar-options), and [line chart](/looker/docs/line-options), among others.\n\nGemini in Looker is a product in the [Gemini for Google Cloud](/gemini/docs/overview) portfolio that provides generative AI-powered assistance to help you analyze and gain valuable insights from your data.\n\nLearn [how and when Gemini\nfor Google Cloud uses your data](/gemini/docs/discover/data-governance).\n\nThis page is intended for users who are exploring data in a Looker instance.\n\nBefore you begin\n----------------\n\nTo create visualizations with Gemini assistance, note the following requirements:\n\n- Gemini in Looker must be enabled for your Looker instance:\n - For Looker (original), the instance must be running Looker 25.2 or later, and the instance must be enabled for Gemini in Looker in the Admin settings. For detailed enablement instructions, see the [Admin settings --- Gemini in Looker](/looker/docs/admin-panel-platform-gil) documentation page.\n - For [Looker (Google Cloud core)](/looker/docs/overview-gemini) the instance must be enabled for Gemini in Looker in the Google Cloud console. For detailed enablement instructions, see the [Administer Gemini on your Looker (Google Cloud core) instance](/looker/docs/looker-core-admin-gemini) documentation page.\n- You must be assigned a Looker role that contains the [`can_override_vis_config`](/looker/docs/admin-panel-users-roles#can_override_vis_config) permission.\n- You must be assigned the Looker [**Gemini**](/looker/docs/admin-panel-users-roles#gemini) role.\n\nUsing the Visualization Assistant\n---------------------------------\n\n| As an early-stage technology, Gemini for Google Cloud\n| products can generate output that seems plausible but is factually incorrect. We recommend that you\n| validate all output from Gemini for Google Cloud products before you use it.\n| For more information, see\n| [Gemini for Google Cloud and responsible AI](/gemini/docs/discover/responsible-ai).\n\nTo customize a visualization with the **Visualization Assistant**, follow these steps:\n\n1. View a supported visualization in an Explore, or edit a visualization in a Look or dashboard.\n2. Open the **Edit** menu in the visualization.\n3. Click pen_spark **Visualization Assistant** to open the prompt menu.\n\nWithin the pen_spark **Visualization Assistant** prompt menu, you can do the following:\n\n- Type out a plain text prompt of changes for the visualization in the **Ask anything** field. To send send_spark your prompt, press return (Mac) or Enter (PC).\n- Apply generated **Suggestions** to the visualization.\n- Open the **Edit Chart Config** to manually write and edit JSON formatting options.\n- **Clear Chart Overrides** to clear all custom changes applied to the visualization. You will be prompted to **Confirm** removing all changes before proceeding. If you cleared the changes by accident it is possible to retrieve them by navigating back in your browser history.\n\nOnce the prompt has run a preview window will open with the options to:\n\n- Manually write and edit the generated JSON formatting options: Click **Edit Chart Config** to open the chart configuration editor. Click **Visualization Assistant** to return to the prompt menu.\n- Start a new prompt: Click **Start Over** to begin a new prompt from the pen_spark **Visualization Assistant** menu. All unapplied changes will be lost.\n- Apply the previewed changes to the visualization: Click **Apply** . If changes are made in the **Chart Config** menu, this option will change to **Revalidate and Apply**.\n- Provide feedback about the helpfulness of the prompt response: Select thumb_up to indicate that the prompt response was helpful, or select thumb_down to indicate that the response was not what was expected.\n\nAn `Invalid JSON detected` error message will appear if you try to preview code containing invalid JSON. You can clean up invalid JSON using the **Autofix code** edit_fix_auto option at the bottom of the **Chart Config** pane.\n| **Caution:** Use caution when editing the default visualization options after making changes with the **Visualization Assistant**. Editing the default visualization options may cause unexpected behavior, including blank visualizations.\n\nCreating successful prompts\n---------------------------\n\nWriting successful prompts can take practice. Try some of the following suggestions to write prompts that successfully accomplish your goal.\n\n### Break prompts into multiple parts\n\nPrompts are more successful when they focus on one change at a time. For example, if you want to change both the font family and the type of visualization, it's better to make these separate prompts. Once you approve of the generated changes apply them and then begin writing the next prompt.\n\n**Prompt**: Change the last series into a line\n\n**Prompt**: Change the font to Arial and make it bold\n\n### Try writing a prompt multiple ways\n\nIf you're not seeing the results that you expect from your prompt, try rewording the prompt. When you apply conditional formatting, try making your prompt as unambiguous as possible. Setting specific value ranges and providing hex codes can create a more successful output. See the following examples:\n\n**Good prompt**: Make the first series #F4B400 if it is over 350000\n\n**Poor prompt**: Highlight the top values in yellow\n\nIf you have multiple series in a visualization, try referring to the series by name rather than by its position in the chart.\n\n**Prompt**: Make the middle series a column\n\n**Better prompt**: Make the users series a column\n\n### Use prompts as a starting point\n\nParticularly complex customizations may require making edits directly to the JSON. Try using the **Visualization Assistant** prompts to create templates and patterns for complex customizations, and then use the **Edit Chart Config** to manually write and edit JSON formatting options. Visit the [Customizing visualizations using the Chart Config Editor](/looker/docs/chart-config-editor) documentation for more on using HighCharts JSON.\n\n### Sample prompts\n\nTry the following sample prompts to get ideas on how the **Visualization Assistant** can help you get started customizing your visualizations:\n\n- \"Make this chart a column chart\"\n- \"Add data labels to the chart\"\n- \"Set the chart margin as 100px\"\n- \"Set the line color as Red\"\n- \"Set the chart background color as pink\"\n- \"Set the line series marker shape as triangle\"\n\nProvide feedback\n----------------\n\nYou can provide feedback for the visualization preview in the **Visualization Assistant** prompt menu. If the generated preview looks correct, click the thumb_up thumbs-up icon. If the generated preview looks incorrect or isn't what you expected, click the thumb_down thumbs-down icon.\n| **Note:** Gemini in Looker is in preview with limited support. We encourage you to share your feedback to help us improve. To report bugs or issues, reach out to [Looker Support](/looker/docs/best-practices/looker-support-details) and be sure to include the following details:\n|\n| - A clear description of the problem and the expected behavior\n| - Steps to reproduce the issue\n| - Any additional relevant details\n\nRelated resources\n-----------------\n\n- [Learn more about Gemini for Google Cloud](/gemini/docs/overview)\n- [Learn more about Gemini in Looker](/looker/docs/overview-gemini)"]]