Piracy: ꜱᴀɪʟ ᴛʜᴇ ʜɪɢʜ ꜱᴇᴀꜱ

Haha. Well, I guess I have a ulterior motive for promoting i2p; the more people get in it and contribute to the network (port forward the NTCP2 and SSU2 to accept transit tunnels) the better network speeds will get! Just the act of torrenting through i2p will help the network tremendously.

I have noticed that speeds in i2pd (c++) are faster than in the original java implementation so maybe you can try that out? It's fast enough for my use case, but yes streaming torrents in i2p is still far away.

But the upside is, you don't have to rely on a VPN to protect yourself against copyright trolls! Plus, putting it on a fully automated *arr stack means you can just add a movie on radarr and let the *arrs take care of everything in about a day or so. I average at about 200-300 KiB / s per well-seeded movie file, and around half that for something poorly seeded.

The speeds have improved tremendously, over the last couple of years some significant improvements have been made. There's still more bandwidth overhead using I2P over a traditional connection but it has been significantly reduced and is not as noticeable anymore. That being said, there's still some configuration that's necessary to maximize your bandwidth. The biggest complaint I hear about the standard i2p install is that it uses extremely conservative bandwidth settings by default but it can all be easily adjusted to maximize performance on your router. I've used I2PD quite a bit but overall I actually much prefer standard Java I2P because it's far more feature rich, more frequently maintained, and settings are muuuuch much easier to configure and understand. There are still many brilliant optimizations in Java I2P that have not made it into I2PD such as the most recent peer analysis techniques that can automatically block/ban misbehaving peers among other things too. I personally think I2PD is best if I just need to host a low resource tunnel.... But back to the speed!

As was already mentioned the more people who participate, the more I2P thrives. One of the most notable differences is that most I2P nodes right now are just enthusiasts running on recycled hardware at a residential address whereas clear net torrents are much more mainstream and many common/popular torrents have at least one peer hosted at a data center with virtually unlimited bandwidth, that one peer usually contributes to over 50% of your download speed on a standard torrent.

I have my router bandwidth setting on my 24/7 router set pretty high and my router usually idles at about 850 KBps... My most recent peak was about 1.3MBps, very acceptable speeds I think. I get the best i2p torrent download speeds using Snark which is built into Java I2P, the only important setting to change is increase tunnel quantity to 10 to maximize your download bandwidth. I have seen some of my downloads seed at about 200+ KBps and I have downloaded at almost a full 1MBps which are comparable speeds to standard clear torrent downloads.

So, in a nutshell, it's not necessarily slower than a standard torrent download (well, maybe a little bit) but what it does have is significantly more variability in bandwidth and download speed depending on how many hops or peers are in between you and your target destination. More hops creates more variance (and more anonymity), you could be directly connected to someone in a data center but the next hop could be connected to a raspberry pi running off of public Wi-Fi which will be the bottleneck in that connection.

You can have qBittorrent running in mixed mode, which doesn't give you the privacy of i2p but does give you even more leachers than just using normal ip and helps grow the i2p network. Everyone should get i2p and use mixed mode or i2p only imo.

Depends on your definition of very slow. I'm currently getting 1MBps which is alright.