Basics

BasicsUpdated Nov 5, 2026
What a context window is, and when you actually need 1M tokens
What counts toward an AI model's context window, how big it is on Claude, GPT-6, Gemini, Grok, DeepSeek and Mistral, and what a 1M-token request costs.
BasicsUpdated Sep 28, 2026
Why does ChatGPT or Claude feel dumber? What's actually on the record
The documented reasons an AI model gets worse: confirmed bugs, changed defaults, outages and long chats. What providers admitted and how to test it.