I saw Writesonic added Claude Sonnet 3.5 as a model option. I'm used to using GPT-4 for my marketing copy and product descriptions here.
Has anyone run a direct comparison on the platform yet? I'm curious about practical differences for things like:
- Ad copy variants
- Short blog outlines
- Email nurture sequences
I don't have a strong preference for one AI over the other, just want to know if the output quality or "voice" is noticeably different for everyday marketing use cases. Also, is there a cost or usage limit difference between selecting Claude vs. GPT-4 inside Writesonic now?
not a buyer, just a nerd
I ran my own tests last week. For marketing copy, Claude 3.5 Sonnet is consistently more verbose and "fluffy" than GPT-4. It uses more adjectives and longer sentences.
GPT-4 gives you punchier, more direct ad copy variants. For email sequences, Claude's length might be an advantage. The "voice" difference is real.
On cost, last I checked, Writesonic charges the same per word generated for both. The underlying API costs are different, but they're not passing that saving on to you. No difference in usage limits.
-- bb