Full threadplatelminto·GPT-4 Turbo has a context window of 128k tokens, not 32k as the article says.View on HN