To evaluate and compare the effectiveness of five (5) widely used AI powered paraphrasing tools such as ChatGPT, Grammarly, Jasper, QuillBot and Copy AI in academic writing, the paper examines their strengths, weaknesses and performance. The research evaluates each of the tools’ ability to enhance lexical diversity, grammatical accuracy, semantic integrity, fluency, and structural modification on a dataset of 50 academic excerpts from diverse disciplines. Our findings demonstrate that while dedicated rephrasing tools exist, QuillBot surpasses the other tools in these types of tasks with the highest achieved lexical diversity (42.3%) and structural modification rate (65.8%) with strong semantic accuracy remaining high (4.6/5). Regarding readability (Flesch score: 58.2), it is close to the leader, and having a very good rephrasing flexibility (60.4% structural changes) makes it a close second. With 1.4 errors per 100 words committed to grammatical correctness (1.4), Grammarly is merely prioritizing correctness (1.4) in 1.4. Though these tools are useful for standard paraphrasing, they are less lexical diverse (31.2% and 21.7%) and less semantically preserved (3.5 and 3.2/5). Based on these results, it would be better to pick paraphrasing tools adequately to address some writing needs: QuillBot for doing in-depth academic rewording, Grammarly for error-free editing, and ChatGPT for balanced paraphrasing and reading.
Academic writing paraphrasing tools AI writing assistants lexical diversity grammatical accuracy text rewording etc
| Primary Language | English |
|---|---|
| Subjects | Autonomous Agents and Multiagent Systems, Artificial Intelligence (Other) |
| Journal Section | Research Article |
| Authors | |
| Early Pub Date | June 9, 2025 |
| Publication Date | June 27, 2025 |
| Submission Date | March 7, 2025 |
| Acceptance Date | May 14, 2025 |
| Published in Issue | Year 2025 Volume: 3 Issue: 1 |