Re:Show HN: Plandex – an AI coding engine for complex tasks
by danenania on Wednesday April 03, 2024 @08:40PM
(#39925180)
Plandex uses gradual summarization as the conversation gets longer (the exact cutoff point in terms of tokens is configurable via `plandex set-model`). So eventually, with a long enough plan, you can start to lose some resolution. That said, assuming you use the default gpt-4-turbo model with a 128k context window, you'd need to go far beyond 2k tokens before you'd start seeing anything like that.
We don't know what ChatGPT's summarization strategy is since it's closed source, but it does seem to be quite a bit more aggressive than Plandex's.
We don't know what ChatGPT's summarization strategy is since it's closed source, but it does seem to be quite a bit more aggressive than Plandex's.