Main Concept
Large Languate Models or LLMs are a specific type of Foundation Model specialized to generate coherent human-like text and are capable to perform language-related task, like translation, summarization, question answering, content creation.
Context
- LLMs are the models behind popular chatbots like ChatGPT, Claude, or Gemini, which have made AI mainstream.
Key Aspects
- LLMs are trained using a large “corpus” of text data, which means a large collections of text documents, like books, articles, websites, newspapers, and other text data
- LLMs are very big models, with billions of parameters
- LLMs are behind not only chatbots but also other type of tools that requires human language type of interaction.
Examples:
LLMs for the main companies in their most current version:
- GPT-5 (OpenAI)
- Sonet 4.5 (Anthropic)
- Gemini 2.5 Pro (Google)
- LLaMA 4 (Meta)
- Amazon Nova (AWS)
- Mistral Large (Mistral AI)
- Cohere Command (Cohere)