This is GSA_ver1_9_00S, a multi-AI assisted scoring application that supports models from OpenAI, Anthropic, and Google.
It supports OpenAI GPT-5.6 (Sol, Terra, and Luna), Anthropic Claude 5 series models, and Google Gemini 3 series models. Users can select among multiple AI models according to their needs, taking into account factors such as response quality, processing speed, and API usage costs.
This application batch-processes students’ written responses and reports organized in Excel files, providing support for scoring, correction, evaluation, and feedback generation. It can also be used to compare the characteristics and scoring results of different AI models under the same grading criteria and prompts.
This application operates through APIs. To use the software, please obtain and configure the necessary API keys from OpenAI, Anthropic, and Google, depending on the AI services you intend to use. API usage is subject to the terms, conditions, and pricing policies of each service provider.
For detailed instructions, please refer to the following article and related materials:
“GPT-ScoreAID: An Application for Supporting the Scoring of and Feedback on Written Responses”
Please note that this application is experimental software developed for research and practical verification of grading support using generative AI. Scores, corrections, evaluations, and comments generated by AI may contain errors, inconsistencies, or variations. Users should always review the generated results and should not rely on them alone for final educational decisions.
Before using this application, please be sure to read the included Terms of Use and Disclaimer and use the software only after agreeing to their contents.