Claude's responses are just word vomit
A user paying for both Claude and GPT $200/month plans is struggling to get Claude to consistently provide concise responses. Despite using models like Opus 4.8, Opus 5, and Fable 5.1, and trying techniques such as the ELI5 trick and adjusting memory, they have not achieved satisfactory results. They are seeking recommendations to resolve Claude's wordiness.
- Published
- 09/07, 21:31 UTC+0
- Ingested
- 09/08, 03:00 UTC+0
- Source type
- Dev community
- Tier
- Community
- Source status
- Healthy
Tier is a per-source editorial setting, not a per-item score.
So I pay for both claude and GPT $200/mo plans. So I get my fair use out of both to say the least. What I cannot figure out is how to get Claude to CONSISTENTLY respond in a manner that is concise and not wordy. I mostly use Opus 4.8, Opus 5, Fable 5.1 I've tried the ELI5 trick, i've tried adjusting memory, etc. Nothing gives me good and consistent results.
I kid you not, I typed "which button should I click next" it gave me a 124 word response. GPT is soooooo nice when it comes to responses, I don't even have to modify the damn thing it just takes my question/task, does it. Then it responds in a beautifully formatted response that has NO FLUFF. A 3,400 word prompt I gave GPT (it was a task), results in a 81 word response. I LOVE THAT.
Both of these models have their place in my business so i'd like to continue using claude.
If anyone has recommendations on how to fix this please let me know.