IPFS for the long term digital preservation: dataset for cost estimate

For the cost estimation described in the study IPFS for the long term digital preservation, and while acknowledging all the caveats associated with this approach, we utilized the capabilities of current large language models (LLMs) to generate project estimates based on the latest publicly available online information. We selected three publicly accessible LLMs: Grok-3, Gemini 2.5 Pro, and DeepSeek-V3. The same prompt was submitted to each LLM, and their outputs were analyzed. We conducted a detailed item-by-item comparison to ensure the responses addressed the requirements specified in the prompt. For each item, we adopted the most conservative estimate, i.e., the highest value provided.In this dataset you can find:● the prompt submitted to the LLMs;● the outputs from the LLMs;● a cost summary derived from the outputs.