ModelRadar
/

NV-Reason-CT Open 3D CT VLM

NVIDIA🇺🇸 USMajor lab· Sep 23, 2026

TextVisionopen weights

NV-Reason-CT is a vision language model for 3D chest and abdominal CT that generates structured reports with step-by-step reasoning and answers follow-up questions. It is an open research foundation for post-training, not a cleared clinical product.

Summary of the model card

Modalities
text, image → text

1 variant

Standard

Sources

Blog
Deutsch

NV-Reason-CT ist ein Vision-Language-Modell für 3D-CT-Aufnahmen von Brust und Bauch. Es erzeugt strukturierte Befunde mit schrittweiser Begründung und beantwortet Rückfragen. Es ist eine offene Forschungsbasis für Post-Training, kein zugelassenes klinisches Produkt.

Provenance

Detected
Sep 27, 2026
First source
labrss
Description
Summary of the model card
Editorially reviewed
—