For the complete documentation index, see llms.txt. This page is also available as Markdown.

Context length limit

LLMs evolve rapidly and you should get the latest models and limitations from official sources

As you may know, each chat model has a different context length limit:

  • GPT-3.5: 4,096 tokens

  • GPT-3.5-16k: 16,385 tokens

  • GPT-4: 8,192 tokens

  • GPT-4-32K: 32,768 tokens

  • Claude: 100,000 tokens

Last updated