Python code for comparing the performance of DeepSeek API, Gemini API, and GPT API on sentiment analysis tasks. This project evaluates how each Large Language Model (LLM) handles sentiment classification using various text inputs.
- Data scraping and preprocessing
- Integration with DeepSeek, Gemini, and GPT APIs
- Sentiment classification using each LLM API
- Comparative analysis of model performance, runtime, and cost
- Confusion matrix visualization for each model
- Python (v3.10 or higher)
- Library on
python-code.ipynb - DeepSeek API (https://platform.deepseek.com/usage)
- Gemini API (https://aistudio.google.com/app/apikey)
- GPT API (https://platform.openai.com/usage)
