For the general purpose section of this course, I would recommend using one of ChatGPT (4o), Claude (3.5 Sonnet) or Gemini (2.0 Flash) based on your personal preference. The models in brackets are the latest and best performing models as of February 2025. If there are newer models available at the time of you reading this, feel free to use those.

Each of these models feel different **to use, and you should test each to see which you interface and response type you prefer. They have slightly different capabilities, pricing models and interfaces.

ChatGPT

ChatGPT is a fantastic all-round platform. The 4o model is great for all tasks, and their o1 model is great for advanced reasoning.

They have the GPT Store, allowing you to use pre-trained bots created for specific purposes and trained on specific data.

They’ve also got the Search functionality, which allows more accurate results from the web with citations included.

You can generate images using DALL-E and upload a variety of files (see types). There is also the Chat GPT Desktop App, which you can download to enjoy a more native experience that includes the ability to take screenshots of your desktop easily.

Pricing

Free / Plus ($20 / month) / Pro ($200 / month)

See a price breakdown here. Additional fees are incurred for additional API usage.

Models

See a list of all available ChatGPT models here.

Claude

Anthropic’s Claude models were the first real competitors to ChatGPT and the performance of 3.5 Sonnet trumps the ChatGPT models in some benchmarks.

The interface is easy to use, and has many of the same features as ChatGPT. You can upload a variety of files (see types),

Pricing

Free / Pro ($20 / month) / Team ($30 / month)

See a price breakdown here.

Models

See a list of all available Claude models here.

Gemini

With the recent release of Gemini’s Flash 2.0 model, Google’s AI capabilities have me the most excited out of all of the models. Accessed via Google’s AI Studio, it offers a clean interface and powerful API.

The Flash 1.5 Pro model can now accept up to 2 million input tokens (see more about Long Context) - which compared to gpt-4o’s 128,000 and Sonnet 3.5’s 200,000 token windows emphasises the leap forward Google has just taken.

This now means you can give Gemini around 10x more input data (in the form of text, images, videos or code). You can ask it to analyse videos up to 1 hour in length, for such things as transcription, scene by scene analysis or audio landmarks.

Pricing

Usage via Google AI Studio is Free in available countries, with pay as you go pricing via the API.

See a price breakdown here.

Models

See a list of all available Gemini models here.

Other Options

Several emerging models deserve attention. Deepseek and Qwen have launched impressive models, showcasing China's growing influence in AI development. Meta's Llama, xAI's Groq, Mistral AI, Minimax, and Amazon's Nova Pro are also making significant strides. The AI landscape continues to evolve rapidly, with new competitors regularly emerging to challenge the current leaders.

Comparing Models

For a comprehensive comparison of model performance, speed, and costs, visit Artificial Analysis.

Based on its free cost, large context window, fast generation speed, and seamless integration with the Google suite, Gemini will be our recommended model throughout this course and will power most of our general-use tasks.

AI for Event Marketing

Modules