ARTICLE PREVIEW

LLMs Comparison - Introduction

Author: Adam Shilton, January 2024

A Comparison of General Large Language Models for Finance: Which One is Best for Specific Use Cases?

The use of Large Language Models (LLMs) has exploded in the past year. They’ve been powering the Generative AI revolution that feels like the industrial revolution all over again.

The most commonly used platform is ChatGPT by OpenAI, but there are others from companies like Google, Meta and Anthropic to name a few.

ChatGPT is not the language model though, it’s merely the platform used to communicate with the OpenAI models.The free version of ChatGPT enables you to communicate with the GPT3.5 model, and the paid version of ChatGPT enables you to communicate with the GPT4.0 model which is more intelligent.

But there is more available to you than just ChatGPT, so we’ve done a comparison of some of the other models available.

What general models did we compare?

ChatGPT-3.5 (Free)
Google Gemini Pro (Free)
Claude Instant (Free)
Llama-2 70b Chat (Free)
CodeLlama-34b Instruct (Free)
Mixtral-8x7B Instruct (Free)
ChatGPT-4 (Paid)
Claude-2.1 200k (Paid)

What platforms did we use?

ChatGPT

ChatGPT is a state-of-the-art language model developed by OpenAI, known for its advanced natural language processing capabilities.

Built on the GPT-3.5 architecture, ChatGPT is designed to understand and generate human-like text in a variety of contexts.

It excels in tasks such as answering questions, engaging in conversation, and providing information on a wide range of topics.

ChatGPT used in following use cases in this research:

To access the free and paid OpenAI language models (GPT3.5 and GPT4.0)

ChatGPT4.0 used in following use cases in this research:

For document and image analysis

Poe

Poe was used to access the other free language models (Gemini Pro, Claude Instant, Llama and Mixtral)

Google Gemini encompasses a suite of multimodal artificial intelligence (AI) large language models proficient in language, audio, code, and video comprehension.

Additionally, Gemini boasts advanced image understanding and recognition capabilities, allowing it to parse intricate visuals like charts and figures without relying on external optical character recognition (OCR) tools.

Google Gemini used in following use cases in this research:

For image analysis

This is a protected part of the content.

Please login below or purchase a membership here to view the entire article.