ArticleBMC oral health2026
Comparison of the ChatGPT and deepseek models in responding to multiple choice questions related to rehabilitation of completely edentulous patients with complete dentures.
Article in BMC oral health, 2026. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Cited by 1 paper.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
1 citing paper in PubMed.
- Article
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
4 authors.
Funding
No grant is acknowledged in the PubMed record.
Abstract
backgroundArtificial intelligence (AI) chatbots are considered a potential resource for dental education. However, there is still a lack of adequate knowledge regarding the trustworthiness, validity, and utility of the content on these platforms, and its application in dental education is still understudied.
methodsA set of Multiple-Choice Questions (MCQs) consisting of 100 questions related to removable complete denture prosthodontics were formulated. Two AI models, DeepSeek-V3 and ChatGPT-4o, were used to deliver the queries. First, accuracy of the generated answers was assessed. Subsequently, two reviewers assessed the LLM (Large Language Model) usefulness and reliability of the responses using modified 5-point Likert scales. McNemar test was done to test the difference between ChatGPT-o4 and DeepSeek-V3 in determining accuracy of scores. The Wilcoxon sign rank test was done to determine the difference between reliability and usefulness between the two AI tools. The Chi square test was used to determine the proportional differences between type of questions and accuracy of scores (α = 0.05).
resultsThe accuracy of the responses was 59% and 66% for ChatGPT-4o and DeepSeek-V3 respectively. There was no significant difference between two AI tools in delivering accuracy of scores (P = 0.281). There was significant difference between reliability scores exhibited by ChatGPT-4o and DeepSeek-V3 (p = 0.027). Deepseek-V3 exhibited a statistically significant higher overall reliability score. Similarly, significant difference in usefulness scores was exhibited in two AI tools (p = < 0.001). Most of the items in ChatGPT-4o showed inaccuracy of responses in analytical based questions than knowledge-based questions and the difference was statistically significant (P = 0.047).
conclusionsThere was no significant difference between ChatGPT-4o and DeepSeek-V3 in delivering accuracy of scores. There are significantly higher reliability and usefulness of responses generated by DeepSeek-V3 compared to ChatGPT-4o. The responses generated by ChatGPT-4o showed greater inaccuracies for analytical-based questions compared to knowledge-based questions.
Indexed as
Identifiers
What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.