Matplotlib font size is one of the most immediate ways to improve how readable and professional your plots look. Setting the right text sizes for titles, labels, and ticks ensures that your visualizations communicate clearly in reports, presentations, and web dashboards.
Below you will find a practical guide to controlling font size in Matplotlib, with examples, best practices, and answers to common questions.
| Parameter | Typical Use Case | Recommended Default | Adjust For |
|---|---|---|---|
| rcParams["font.size"] | Base text size for labels and annotations | 10–12 | Overall document readability |
| rcParams["axes.titlesize"] | Figure and subplot titles | 14–16 | Emphasis and hierarchy |
| rcParams["axes.labelsize"] | Axis X and Y labels | 12–14 | Clarity of axis meaning |
| rcParams["xtick.labelsize"] & rcParams["ytick.labelsize"] | Tick labels on each axis | 10–12 | Dense plots or small figures |
| rcParams["legend.fontsize"] | Legend text | 10–12 or "medium" | When legends contain many entries |
| rcParams["figure.titlesize"] | Figure-level titles (Figure suptitle) | 16–18 | Multi-plot dashboards |
Setting Global Font Size with rcParams
The quickest way to control Matplotlib font size across an entire project is to update rcParams before creating any figures. By setting a base size and axis-specific overrides, you keep consistent scaling without repeating arguments in every function call.
Use matplotlib.rcParams.update with a dictionary or assign individual keys to adjust titles, labels, ticks, and legends globally. These settings apply to all subsequent plots in the same Python session or script.
Font Size Per Text Element in Plotting Calls
Title and Label Arguments
Many plotting functions accept size-related keyword arguments such as fontsize for set_xlabel, set_ylabel, and set_title. Supplying these values locally is useful when each subplot needs a different scale.
Tick and Legend Size Options
You can pass fontsize to tick_params for axis ticks and to legend for the legend text. This approach keeps the rest of the figure at the global defaults while fine-tuning dense or specialized visuals.
Using Font Manager and LaTeX Fonts
When precise typography matters, you can pair Matplotlib font size settings with custom font families managed by the font_manager. Specify family names in rcParams["font.family"] and size in rcParams["font.size"].
If you use LaTeX rendering, adjust text.usetex in rcParams and set compatible sizes to ensure that symbols and Greek letters remain crisp at any resolution. Larger sizes often improve legibility in exported PDF figures.
Responsive Sizing for Different Output Formats
Figure dimensions and DPI interact strongly with Matplotlib font size. A plot that looks clear on screen may have cramped or oversized text when saved as PNG or PDF. Scale font elements proportionally when changing figsize or dpi.
For publication-quality output, increase font sizes and test vector formats like SVG or PDF. For slide decks, slightly bolder sizes improve visibility without forcing you to magnify the entire figure.
Best Practices for Matplotlib Font Size
- Define a consistent base size in rcParams and only deviate when emphasizing hierarchy.
- Test figures at the final output size, not just on screen, to catch readability issues.
- Increase tick and label sizes when using high DPI exports for reports or posters.
- Keep legend and annotation sizes aligned with axis label sizes for visual balance.
- Automate adjustments with a small configuration block at the start of your notebook or script.
FAQ
Reader questions
How do I change only the axis label sizes without touching other text?
Use ax.set_xlabel and ax.set_ylabel with the fontsize argument, or update rcParams["axes.labelsize"] to affect all future axis labels uniformly.
Why do my tick labels appear too small after I change the figure size?
When you enlarge or shrink figsize, the physical space of ticks does not automatically scale. Adjust rcParams["xtick.labelsize"] and rcParams["ytick.labelsize"] or pass fontsize to tick_params to restore readability.
Can I set different font sizes for each subplot in a figure?
Yes. Access each subplot axis with ax.set_title, ax.set_xlabel, and tick_params individually, providing a specific fontsize value per subplot.
Does changing font size affect layout or overlapping text in tight plots?
Larger text increases label and title height, which can cause clipping or overlap in constrained layouts. Use plt.tight_layout or adjust subplot parameters to prevent this.