LLMs: large language models

LLMs (Large Language Models) are the engine of the AI wave. This dossier explains what they are, how to compare them, and tracks their releases — from Claude to GPT, Gemini, Mistral and Llama.

Latest LLM news

What is an LLM?

A large language model is a neural network trained on vast text corpora to predict and generate language. From this capability emerge writing, reasoning, code and tool use.

How to compare them

Beyond benchmarks, what matters: writing quality, reasoning, context window, speed, cost, and openness (proprietary vs open weights). See our detailed comparisons.

The current landscape

Claude (Anthropic), GPT (OpenAI) and Gemini (Google) lead on the proprietary side; Mistral, Llama, DeepSeek and Qwen push open source. Each has its strengths.

Frequently asked questions

What does LLM mean?

Large Language Model — trained to generate and understand text.

Is Claude an LLM?

Yes, Claude is a large language model built by Anthropic.

What's the best LLM?

It depends on use: see our comparisons (writing, code, cost, context) to decide.

← Claude news in real time

Claude News is published by Héra SASU. Independent media, not affiliated with Anthropic.