The benchmark, developed with dentists, dental hygienists and dental assistants, pairs a written charting standard with test scripts, expected charts and a score card. Its rules cover Universal tooth numbering, six sites per tooth, reading order, missing teeth, corrections, findings and scope. To pass a test, software must place every expected entry at the correct tooth and site without producing extras.

The eight levels range from explicit tooth-and-site readings to complete examinations. They test automatic navigation, periodontal findings such as bleeding and recession, flexible phrasing, simulated room conversation and corrections, shorthand dictation, safety rules and consistency between full-word and shorthand exams. DentScribe reported perfect results across all 24 tested dictation styles.

The company also said the system passed 13 additional noisy-room text scripts, with 471 of 471 expected entries correct and no extra entries. Those scripts include simulated speech-recognition mishearings; they are not a live test of suction noise or microphones. Across all 81 clean and noisy scripts, eight runs per script produced identical charts.

For the second complete exam, the shorthand script used 210 words, compared with 397 in the full-word version—about 47% fewer. DentScribe cautioned that this measures script length, not chairside time saved. The benchmark kit, including read-aloud scripts, expected charts and a score card, is available from the company for practices to evaluate charting systems.