Chart-To-Code Generation Benchmark For Multimodal Models

ChartMimic evaluates visually-grounded code generation in large multimodal models using information-intensive visual charts. The dataset was created by the ChartMimic team and was last updated in June 2025.

Use Cases

• Benchmarking LMMs on generating code from visual chart elements.
• Evaluating cross-modal reasoning by assessing code output for chart data extraction.
• Training models to synthesize executable code based on chart images and textual descriptions.

Strengths

• Designed for a specific evaluation task: chart-to-code generation.
• Integrated into the VLMEvalKit framework for standardized assessment.

Limitations

• Specific row count, column names, and dataset size are unknown.
• Data collection methodology and temporal coverage are unspecified.

Provenance

Source

ChartMimic

Collection Method

null

Time Range

null

Freshness

Last updated on 2025-06-17.

Geography

null