diff --git a/firmware/tools/edit-font.py b/firmware/tools/edit-font.py
new file mode 100644
index 000000000..9e03467ae
--- /dev/null
+++ b/firmware/tools/edit-font.py
@@ -0,0 +1,107 @@
+# Tool to extract font data, save to bitmap file, and apply modifications
+# to source code after editing the bitmap.
+
+from __future__ import print_function
+from matplotlib.pyplot import imread, imsave, imshow, show
+import numpy as np
+import sys
+import re
+
+if len(sys.argv) != 3:
+    print("Usage: %s <font name> { show | extract | insert}" % sys.argv[0])
+    sys.exit(1)
+
+font_name = sys.argv[1]
+mode = sys.argv[2]
+
+# Where to find the font data - may need updated if code has changed.
+source_file = 'firmware/common/ui_portapack.c'
+data_marker = f'static const uint8_t {font_name}_glyph_data[] = {{'
+data_offset = 1  # Data starts this number of lines after start marker.
+end_pattern = '(.*)};$' # Indicates end of font data.
+def_marker = f'static const ui_font_t {font_name} ='
+def_offset = 1
+def_pattern = '^\t{{(?P<width>\\d+), (?P<height>\\d+)}'
+
+# Image filename used in extract and insert modes.
+image_file = f'{font_name}.png'
+
+# Get the literals used to populate the font array in the source file
+lines = [line.rstrip() for line in open(source_file).readlines()]
+try:
+    start = lines.index(data_marker) + data_offset
+except ValueError:
+    print(f"Could not find glyph data for font '{font_name}'")
+    sys.exit(1)
+
+data = str()
+for line in lines[start:]:
+    if match := re.match(end_pattern, line):
+        data += match.group(1)
+        break
+    else:
+        data += line
+
+# Get the width and height from the font definition.
+try:
+    def_start = lines.index(def_marker)
+except ValueError:
+    print(f"Could not find definition for font '{font_name}'")
+    sys.exit(1)
+
+def_match = re.match(def_pattern, lines[def_start + def_offset])
+
+if def_match is None:
+    print(f"Failed to parse definition of '{font_name}'")
+
+width = int(def_match.group('width'))
+height = int(def_match.group('height'))
+
+# Eval the literals to get the values into a numpy array
+packed = eval("np.array([%s], np.uint8)" % data)
+
+# Reorganise into a monochrome image.
+chars_per_row = 5
+unpacked = np.unpackbits(packed)
+bitmaps = (unpacked
+    .reshape(-1, height, width // 8, 8)[:,:,::-1,:]
+    .reshape(-1, height, width)[:,:,::-1]
+)
+indices = np.arange(len(bitmaps)).reshape(-1, chars_per_row)
+image = np.block([[bitmaps[idx] for idx in row] for row in indices])
+
+if mode == 'show':
+    # Display font image
+    imshow(image, cmap='binary')
+    show()
+
+elif mode == 'extract':
+    # Save font image
+    imsave(image_file, image, format='png', cmap='binary')
+
+elif mode == 'insert':
+    # Read in modified font image
+    image = imread(image_file)[:,:,0].astype(bool)
+
+    # Reorganise back to original order
+    rows = image.reshape(-1, height, image.shape[1])
+    single_row_image = np.hstack(rows)
+    num_chars = single_row_image.shape[1] // width
+    bitmaps = np.array(np.split(single_row_image, num_chars, 1))
+    unpacked = (bitmaps
+        [:,:,::-1].reshape(-1, height, width // 8, 8)
+        [:,:,::-1,:].reshape(-1)
+    )
+    packed = ~np.packbits(unpacked)
+
+    # Replace lines of file in same format as used before
+    bytes_per_row = 13
+    split_indices = np.arange(bytes_per_row, len(packed), bytes_per_row)
+    for i, group in enumerate(np.split(packed, split_indices)):
+        line = ("\t0x%02x," + " 0x%02x," * (len(group) - 1)) % tuple(group)
+        if len(group) < bytes_per_row:
+            line = line[:-1] + "};"
+        lines[start + i] = line
+
+    # Write out modified source file
+    open(source_file, 'w').writelines([line + "\n" for line in lines])