My specific needs include:
- Token counting and cost estimation tools.
- Visibility into the sliding window of context in chat-style environments, as it's unclear which parts are considered in OpenAI's UI.
- Code compression and optimization techniques, such as stripping comments, removing irrelevant utility functions, or other methods often discussed here.
- Assistance with frontend coding: For instance, I struggle with managing React projects that have numerous files and require multiple changes for data passing. While the 32k model might help, it can be expensive.