Prompt Optimization Strategies for Large Language Models: Insights from Random Prompting with Gemini API
International Journal of Research in Computing · 16.07.2026 · 43
DOI: 10.64701/ijrc/345/9123 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Research in Computing, online erschienen am 16.07.2026. Verfasst von OAM Jayawardana.
Abstract (Originalsprache)
The increasing use of large language models (LLMs) has raised concerns about efficiency regarding token usage, as excessive token consumption burdens computational resources. This study addresses the need for systematic techniques to optimize prompts, evaluating three specific strategies within the Gemini API framework aimed at reducing token usage. The research examined 480 prompts across 16 models in the Gemini family to quantify potential token savings. The three strategies assessed are structured concise formatting, which utilizes bullet lists, headings, and concise language to streamline content; verbose removal, which eliminates redundant and irrelevant information to enhance clarity and focus; and code formatting, which promotes consistent syntax while minimizing inline commentary. Token counts were tracked before and after optimization using the Gemini countTokens API endpoint. Findings revealed an average token saving of 37.15% across most Gemini-family models, with Gemma 3 models achieving a higher
Zitieren
APA
Jayawardana, O. (2026). Prompt Optimization Strategies for Large Language Models: Insights from Random Prompting with Gemini API. International Journal of Research in Computing, 43. https://doi.org/10.64701/ijrc/345/9123
Vancouver
Jayawardana O. Prompt Optimization Strategies for Large Language Models: Insights from Random Prompting with Gemini API. International Journal of Research in Computing. 2026:43. doi:10.64701/ijrc/345/9123
BibTeX
@article{Jayawardana2026,
author = {Jayawardana, OAM},
title = {Prompt Optimization Strategies for Large Language Models: Insights from Random Prompting with Gemini API},
journal = {International Journal of Research in Computing},
year = {2026},
pages = {43},
doi = {10.64701/ijrc/345/9123}
}
EUVIMED