We’re excited to announce Unovis 1.7.0, our largest release to date.
This release introduces new capabilities for developers building AI-powered applications, server-rendered experiences, and secure production environments. It also adds three new chart components and includes improvements to accessibility, performance, and rendering reliability.
What’s new in Unovis 1.7.0
@unovis/mcp— Generate Unovis charts through an AI agent@unovis/ssr— Render charts server-side in Node.js- Three new chart components: boxplot, heatmap, and radial chart
- Improved performance across the library
- Enhanced color and pattern support for more WCAG-compliant charts
- Support for strict Content Security Policies
- Standalone SVG output for server-side rendering workflows
AI-powered chart generation with @unovis/mcp
Unovis 1.7.0 introduces an MCP server that allows an AI agent to generate charts from data.
The new @unovis/mcp package can transform data into Unovis charts and return several output formats:
- SVG
- PNG
- A live HTML file
- Ready-to-paste component source
The rendering pipeline runs locally in Node.js. It does not require a browser or a remote rendering service, and the data does not leave the local environment.
This makes @unovis/mcp useful for applications such as:
- AI-assisted data exploration
- Natural-language chart generation
- Developer tools that produce chart source code
- Internal analytics workflows
- Automated chart and report generation
- Environments where data privacy is a requirement
The generated component source can also provide a starting point for developers who want to refine the resulting chart manually.
Server-side rendering with @unovis/ssr
Unovis now provides a dedicated package for headless, server-side chart rendering:
npm install @unovis/ssr
The @unovis/ssr package provides a jsdom-based rendering environment with the infrastructure needed to render charts outside of a browser, including:
- Canvas text metrics
- Geometry polyfills
- A flushable animation-frame queue
- Post-processing for standalone SVG output
The post-processing step produces SVG that does not depend on:
- CSS classes
- CSS custom properties
- External references
This makes the output easier to embed in server-rendered applications, generated documents, static assets, and environments where a browser is unavailable.
The package is also useful for applications that need to render charts during a server-side request, generate chart assets ahead of time, or produce portable SVG output for downstream systems.
New chart components
Unovis 1.7.0 adds three new chart components:
Boxplot
The boxplot component supports statistical distribution analysis, including the visualization of quartiles, medians, ranges, and outliers.
Heatmap
The heatmap component supports dense, matrix-based data exploration and is useful for displaying patterns across two dimensions.
Radial chart
The radial chart supports circular data representations and hierarchical or part-to-whole visualizations.
These additions expand the chart types available to developers while preserving Unovis’s composable approach to chart construction and configuration.
Content Security Policy support
Unovis now supports strict style-src Content Security Policies.
This is important for applications that restrict inline styles or operate under more stringent security requirements. The improvement makes it easier to integrate Unovis into production applications with an established Content Security Policy rather than requiring developers to weaken those policies for chart rendering.
Accessibility and styling improvements
Unovis 1.7.0 includes improvements to colors and patterns that help developers create more accessible charts and improve visual differentiation.
These changes support more WCAG-compliant chart configurations, particularly in scenarios where color should not be the only way to distinguish data series or categories.
As always, developers should validate the final chart configuration against the accessibility requirements of their application, including contrast, labeling, keyboard interaction, and alternative representations where appropriate.
Performance improvements
This release also includes broad performance improvements across the library.
These improvements are especially relevant for applications that render multiple charts, work with larger datasets, or combine client-side rendering with server-side generation.
Project growth
Unovis now has approximately 216K weekly downloads, up from 43K one year ago—a nearly fivefold increase.
For complete details, implementation guidance, and release information, see the official release notes:
We’d love to hear how you’re using these new capabilities. If you have technical questions, feedback, suggestions, or ideas for future releases, please reach out to the Unovis team or the UI Foundational team. We’re always happy to chat.
You can contact:
- Qian Liu — qi.liu@f5.com
- Surya Teja Hanumandla — s.hanumandla@f5.com