Character Error Rate

A metric for evaluating automatic speech recognition (ASR) and optical character recognition (OCR) accuracy, measuring the minimum number of character-level edits (insertions, deletions, substitutions) needed to transform the system output into the reference text, divided by the total number of characters in the reference. CER is particularly useful for languages that do not separate words with spaces (such as Chinese or Japanese) and for evaluating fine-grained accuracy where individual character errors matter. Compared to Word Error Rate (WER), CER is typically lower numerically but may not directly correlate with comprehensibility—a single character error can change meaning entirely (e.g., "not" vs "now").