Medmio’s CodeSight AI medical coding engine reached 98.1% accuracy on charts it had never seen — published with the complete evidence base behind every number.
SAN DIEGO, CA, UNITED STATES, July 22, 2026 /EINPresswire.com/ — Medmio has released the Medmio EM-945 benchmark, a public, independently checkable measure of how accurately its AI reads a clinical note and assigns the billing codes. On charts it had never seen, Medmio’s CodeSight(TM) engine matched the exact billing level a vascular practice actually billed 98.1% of the time (95% confidence interval 96.2-99.1%), with no custom training on the practice’s data — and any chart it was not confident about was routed to a human coder for review.
CodeSight is the coding engine inside Medmio’s broader platform — which also handles voice dictation, patient intake, and mobile charge capture — and pairs an ensemble of large language models with deterministic, policy-driven rules so its output is both accurate and consistent.
The benchmark arrives as many practices, short on experienced medical coders, weigh how far to trust automation with their billing.
Medmio published the full basis for the 98.1%: the confidence intervals, a calibration analysis, and the complete accuracy-coverage curve — all drawn from 945 billed-and-paid charts, scored against the codes the practice actually billed, on a test set locked before the result was measured.
“We are very pleased to publish this benchmark,” said Levan Sulakvelidze, President and CEO of Medmio. “98.1% is a great initial result — but what excites me most is that we reached it with no custom training on this practice’s data, a clear indication CodeSight can deliver this level of quality, and higher, for practices everywhere. There is a real opportunity here to give physicians their time back and their patients more of their attention.”
The benchmark covers office visits at one practice; Medmio says it will publish future results the same way, on larger datasets and a broader, more complex range of codes. The company frames that openness as a standard it means to hold its own technology to — proof it believes healthcare AI should offer as a rule, not an exception.
The full report, with all tables and figures, is available with a PDF download here: CodeSight Benchmark
About Medmio
Medmio is an AI-powered platform for physician practices that brings autonomous medical coding, voice dictation, patient intake, and mobile charge capture into one system. Its CodeSight(TM) engine reads clinical notes and produces billing and diagnosis codes with reviewable confidence scores, and integrates with major EHRs via HL7 standards. Learn more at www.medmio.com.
Media Contact
Levan Sulakvelidze, President & CEO
Medmio, Inc.
press@medmio.com
https://www.medmio.com/press-kit/
Levan Sulakvelidze
Medmio, Inc.
lsulakvelidze@medmio.com
Visit us on social media:
LinkedIn
Instagram
Facebook
YouTube
X
Legal Disclaimer:
EIN Presswire provides this news content “as is” without warranty of any kind. We do not accept any responsibility or liability
for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this
article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
Media gallery
