The rising cost, time burden, and ethical constraints associated with collecting human behavioral data create persistent challenges for researchers who design, pilot, and test behavioral theories. These challenges are especially important in marketing, psychology, information systems, management, and social sciences, where experiments often require large samples, multiple conditions, repeated pretests, or sensitive manipulations. Large language models (LLMs) offer a promising complementary approach by generating synthetic participant responses that can help researchers evaluate stimuli, refine survey instruments, explore theoretical predictions, and compare expected response patterns before costly human-subject data collection. However, using LLMs for this purpose often requires substantial technical expertise, including model selection, API configuration, prompt engineering, persona construction, response parsing, and reproducible data management. SEDG (Synthetic Experimental Data Generator) is a software platform that lowers these barriers by enabling researchers to generate synthetic responses for behavioral experiments using local and cloud-based LLMs. The platform is available at https://synthstudy.vercel.app/ and is designed around the workflow of experimental researchers rather than software engineers. SEDG allows users to specify study stimuli, define survey questions, select LLM providers and models, configure generation parameters, use predefined personas, provide custom personas, or generate responses without personas. It supports diverse experimental designs and produces structured outputs for analysis, replication, and comparison across models and conditions. By integrating flexible LLM access with research-oriented prompt construction and export functionality, SEDG democratizes access to LLM-based behavioral simulation across disciplines.