This module covers practical techniques for effectively using Large Language Models in various applications and contexts.
| Name | Description |
|---|---|
| ChatGPT | OpenAI’s flagship conversational AI featuring GPT-5.1, GPT-o3-mini, and o1 reasoning models. Excels at natural conversation, coding, analysis, multimodal tasks, and complex reasoning. Image generation available for all users. Pro and Team tiers unlock enhanced voice, workspace tools, and canvas collaboration. Strong general-purpose assistant. |
| Claude | Anthropic’s assistant focused on safety and advanced reasoning. Powered by Claude 3.7 Sonnet and Claude 3.5 Opus, with ~200K–1M token context. Great for structured writing, code generation (especially with Claude Code), document analysis, and careful reasoning. “Constitutional AI” safety design and increasingly capable multimodal features. It’s one of my favourite for code generation, especially using the Claude Code agent |
| Perplexity | AI-powered research and search assistant combining strong LLMs with live web access, citations, and retrieval. Ideal for fact-checking, scientific queries, news, and exploratory research. Pro subscription provides access to GPT-5, Claude, and custom models with larger context windows and file-upload workflows. |
| Google Gemini | Google’s multimodal AI with Gemini 2.5 Pro/Flash/Ultra. Integrated across Google Workspace with 1M+ token context window. Best for productivity tasks, document processing, coding, and Google ecosystem integration. Features include native image generation, deep research capabilities, and enterprise-grade security. |
| Mistral AI | French AI company offering efficient, open-source models including Mistral Large 2 and Pixtral. Features 128K context windows and strong multilingual support. Best for European data compliance, code generation, and cost-effective deployment. Emphasizes transparency and developer-friendly licensing. |
| Meta LLaMA | Meta’s fully open-weight models (e.g., Llama 4, Llama 3.1, Scout, Maverick) with large context windows and mixture-of-experts architecture. Optimized for customization, fine-tuning, and research. Popular in academia and industry due to permissive licensing and strong performance relative to proprietary models. |
| Microsoft Copilot | Microsoft’s AI assistant integrated across Office 365, Windows, and web. Uses GPT-5-based models and Microsoft’s own Phi-3 family for on-device tasks. Recommended for productivity workflows, enterprise environments, and Microsoft ecosystem users. Features include document assistance, meeting summaries, and code generation. |
| DeepSeek | Chinese AI company offering cost-effective models with strong reasoning capabilities. DeepSeek R1 features competitive performance at lower costs ($0.55/$2.19 per 1M tokens). Not recommended online, but the model can be ran locally. |