Blind children are individuals who experience abnormalities in the sense of vision, so that the eyes cannot function normally. This child is one type of child with special needs who has abnormalities in the sense of vision that require services, both in terms of education and health specifically. To overcome this, media is used as a tool used for delivery, one of which is a Sensor-Based Education Model through Phantom 3D (PHANKESGI) Against Tooth Brushing Skills. Research Purpose: Produce Sensor-Based Education Model through Phantom 3D (PHANKESGI) as a learning medium to improve toothbrushing skills and reduce debris index in blind children. The method used in this research is Research and Development (R & D) with 5 stages of research, namely information collection, model design, expert validation and revision, model / product trials and model / product results. The quasy experiment design used is a non randomized pre and post-test group control design. Purposive sampling technique, consisting of 5 intervention group children and 5 control group children, with treatment duration given for 21 days. The results of 3D expert validation (PHANKESGI) showed a p-value of 0.858 and an average of 90.8%, which means that this model is feasible as an effort to improve toothbrushing skills in blind children. This model was effective in improving knowledge (p = 0.001), attitude (p = 0.001), and action (p = 0.001) towards teachers and effectively improving brushing skills (p = 0.005) and lowering debris index scores (p = 0.041) compared to the control group.Conclusion: 3D Development (PHANKESGI) is effective in improving toothbrushing skills in blind children.