Which LLM Should You Choose?

A guide for choosing the best LLM API for your use case

eneral Recommendation

Use Case	Use Case Description	Preferred Model(s)
General Tasks (Open-Ended)	Handle diverse requests without specific instructions. Relies on the model's broad knowledge and understanding.	GPT-4-Turbo
General Tasks (Guided)	Handling a wide range of tasks with the help of few-shot or chain-of-thought prompting, enabling smaller, faster models to produce reasonably good quality outputs with reduced latency and cost.	Claude 3 Haiku
Low Latency & Cost-Sensitive	Prioritize quick responses and affordability over the absolute best output quality. Ideal for real-time or high-volume applications.	Claude 3 Haiku/Mixtral
Auto Prompt Rewriter and Refinement	Automatically refine prompts for better results. Understands intent and suggests changes for improved effectiveness with smaller LLMs.	GPT-4-Turbo/ Claude 3 Opus

Core Applications:

Use Case	Model
Language Understanding and Generation: Content Creation and Summarization Language Translation and Localization Conversational AI and Virtual Assistants (Chatbot) Sentiment Analysis Intelligent Writing Assistance	Claude 3 Haiku Claude 3 Opus for Multilingual
Computer Vision and Multimodal Integration: Image Captioning Visual Question Answering Multimodal Data Processing for Enhanced Insights	Claude 3 Haiku
Information Management: Data Retrieval and Analysis Knowledge Management and Research Support Personalization and Recommendation Systems Visual/Text Search	Option: Retrieval-augmented generation (RAG) Model: Claude 3 Haiku/ Command R+

Specialized Applications:

Use Case	Model
Code and Automation: Development and Coding Support: Automating code generation and offering software development assistance. Operational Efficiency: Enhancing process automation and intelligent document processing for improved workflow. Content and Security Management: Implementing content moderation, filtering, and mechanisms for fraud detection and prevention.	Claude 3 Opus
Business and Commerce: E-commerce and Retail Enhancements: Optimizing online shopping experiences and supporting retail operations. Customer Engagement: Utilizing chatbots for service and support, and deploying personalized email marketing campaigns. Marketing and Sales: Developing targeted marketing strategies, generating product descriptions, and leveraging intelligent upselling and cross-selling. Market Insights: Conducting market research, analyzing customer reviews, and forecasting demand. Product Management: Automating product categorization, tagging, and inventory management.	Claude 3 Haiku
Education and Learning: Personalized Education: Creating adaptive learning materials and tools tailored to individual learning styles and needs.	Claude 3 Opus
Geospatial and Environmental Analysis: Spatial Data Insights: Offering mapping and spatial data analysis for various applications.	Claude-3 Sonnet

Performance vs Price Analysis

Input token price: This graph illustrates a comparison of popular large language models (LLMs), plotting 'LMSYS Better Arena Elo' against 'Price per 1M Token (Input)', both on a logarithmic scale.