//
Search across all documentation pages
10 pages in this section.
Learn how Claude processes your prompts, from tokenization and context window assembly to reply generation. Understand key concepts and practical limits.
Learn how tokens, context windows, and turn-based chat structure work in Claude conversations. Explore examples of sending messages and tokenization.
Learn how Claude converts text into tokens, why token counts differ from word counts, and how tokenization impacts pricing and usage limits.
Learn how system prompts, user turns, and assistant turns structure Claude conversations, defining behavior and responses throughout a chat.
Learn how Claude's session memory works, why it "forgets" across chats, and the one option for sharing context across separate conversations.
Learn best practices for managing tokens, context limits, and multi-turn conversations with Claude to improve effectiveness and fluency.
A single-page roundup of every highlight bullet from the 9 pages in the Context & Tokens section, grouped by source page so you can scan all 38 takeaways without opening each article individually.