Is "min first post typing time" still available?

AI-generated summary

Window_Plants reports that after updating Discourse, the previously effective spam-prevention setting “min first post typing time” is no longer available, replaced by “Fast typing threshold” set at 5 seconds on “high” — which they find too short, leading to renewed spam. They request the return of the original setting with custom value support. sam asks if they’re using AI spam detection, noting that relying solely on typing time may cause legitimate users to be incorrectly flagged. Window_Plants confirms they haven’t used AI detection and preferred the simplicity of the old setting. sam suggests trying AI spam detection (with free Gemini API keys) and mentions that typing for 5–10 seconds is normal for many users. As an alternative for low-volume forums, they recommend placing all TL1-0 users in the approval queue. The thread highlights a tension between spam control and user experience, with the community seeking a return of a straightforward, customizable anti-spam tool.

Hello folks!
My forum has been hit by spam bots a few months ago and setting the “min first post typing time” to a higher value worked wonders.

I recently updated Discourse and can’t find that setting anymore. It seems to have been replaced by “Fast typing threshold” which is 5 seconds at the “high” setting. This unfortunately is too short and I got inundated by spams this morning :sweat_smile:
There’s no option to put a custom value either.

Please let me know how I can get access to “min first post typing time” with custom value option again or if it’s still around and that I have been looking at the wrong place.

Thanks so much!

Can you confirm if you are using AI spam detection? Are you relying solely on fast typing threshold for all spam detection?

Hi Sam! Thank you for your response :slight_smile:

I haven’t been using the AI spam detection.
I was solely using the “min first post typing time” to prevent the spams, which was working fine for me.

Would it be possible to have that feature back in the next update? That would be amazing as it was simple and straightforward to use.

I think we can consider it, but can you please try out ai spam detection first? You can get free scanning using Gemini free api keys so it will only cost you time to configure.

We have seen other communities reach very large backlogs due to users entering the queue incorrectly. Typing for 5-10 seconds is perfectly normal in many cases. “Yes, this was very helpful”

Another alternative for a low volume tightly moderated forum is just to put all tl1-0 in the queue.