I think you might be referring to Claude 2 (100K context) or the 100K token context window that Anthropic introduced.
Here's what you likely mean:
Claude 2 (released July 2023) was the first model Anthropic released with a 100,000 token context window. This was a big deal at the time because:
- It could process roughly 75,000 words in a single prompt
- This equates to about 300-500 pages of text
- It could handle entire novels (e.g., The Great Gatsby fits easily, even longer books)
- It allowed users to upload very long documents and ask questions about them
What made 100K context special?
- Previously, most models had ~4K-32K context
- 100K meant you could have extended conversations without losing early context
- You could analyze entire codebases, legal documents, or research papers in one go
Current models:
- Claude 3.5 Sonnet and Claude 3 Opus now support 200K tokens
- Claude 3 Haiku supports 200K tokens as well
- So 100K was a milestone, but current models have double that
Were you asking about something specific regarding Claude's 100K context window, or did you have another question in mind?
