Thoughtpile I write things here.

My first impressions with Misskey

I switched from Pleroma to Misskey about a week ago. I’m writing down my first impressions here. I run a single user instance, so i can’t tell you anything about the local community functions. Keep in mind that these are personal experiences, i left a lot out and i might have misunderstood some things.

The federation is working better

Pleroma had problems federating with BookWyrm and PeerTube, i couldn’t receive posts from these platforms. It works flawlessly with Misskey.

Simpler configuration

Misskey’s configuration file is very short and contains only the absolutely necessary. Everything else is configured in the web interface, although the documentation can be improved in some areas. Pleroma on the other hand seems to be much more flexible, at the cost of a more complicated configuration.[1]

Custom emojis

Pleroma has them too, but Misskey can send them as reaction (Mastodon and BookWyrm show a favourite instead, other software, like Pleroma, ignores it). The emojis of remote instances can be searched and imported.

MarkDown support

Like Pleroma, Misskey supports MarkDown when composing messages. However, there are some differences. Misskey doesn’t support lists and headlines at the moment, but has some fun extra features, like a rainbow effect, bouncing text or a sparkling effect. I recommend trying $[shake $[bounce $[x4 $[rainbow $[sparkle Ü]]]]]. 😉

Suggestions for hashtags

I always missed that. When you write a hashtag, you get suggestions based on the ones you used before and others your server knows about.

Resource usage, server

My Misskey instance with 1 user uses a little bit less than 600 MB RAM at the moment. With Pleroma it was a bit less than 300 MB. On top of that there are about 100 MB for PostgreSQL for both of them. I don’t know how that scales with more users. The CPU usage seems to be negligible for both.

Resource usage, web interface

The web interface of Misskey is considerably more CPU intensive, particularly when you load it first. The RAM usage seems to be about the same, a little less than 20 MB according to about:performance in Firefox. On my more than 10 year old laptop with Firefox without hardware acceleration it takes about 13 seconds to load. After that the experience is smooth.


They work approximately like watching hashtags, but they can look for normal text as well. You can add as many phrases as you like and the results can be filtered with an ignore list.


There is a Deepl integration! But in order to use the gratis API you have to give them your credit card details. 🙁 The results are mixed, but often good enough.

Thread view

A lot better in Misskey, the levels are visualised properly, like in my e-mail program. 💖

Follower import

I forgot how it was when i switched to Pleroma. When i switched to Misskey, i exported the list in Pleroma and imported it in Misskey, done. Or so i thought. It seems the list was too long, or maybe something else happened. In any case, the import stopped around the 100th account. I worked around that by creating a new list from the differences of the Pleroma export and the Misskey export, with the help of the Linux command line: comm -23 pleroma.csv misskey.csv > importthis.csv (sort them with sort beforehand).


For replies, the note (post, toot) that is answered to is also shown. Nice feature for renotes (repeats, boosts), for long threads from people i follow, not so much.


The API is not compatible with Mastodon’s, so programs written for Mastodon will not work. SubwayTooter works pretty well, but i don’t like the user interface. Milktea looks nice but it is in an early development stage. At the moment, i use the PWA on my pocket computer. It sucks a lot of power.


A programming language that apparently can be used to write plug-ins and scripts. I looked at it briefly,[2] but didn’t engage with it seriously. Looks exciting!


Collections you can add notes to. They can be public or not. Useful to store favourites (bookmarks) in categories or to collect scattered wisdom into one place.

1. I used Pleroma as a ‘from source’ installation. ‘OTP’ installations seem to be easier to configure.
2. My insights so far are collected here: