aptea on Nostr: Proton's AI service, Lumo, has really come a long way since it first launched. When ...
Proton's AI service, Lumo, has really come a long way since it first launched. When it first came out, I found it relatively useless compared to the big tech LLMs and even Venice AI (another privacy-respecting AI service), which also was lagging behind at that point.
Now, Lumo is my go-to in most cases unless I'm using AI to do something that I want to save long-term (query history is retained for only 7 days on the free plan) or if I need a more advanced model, in which case I tend to use Venice (or Perplexity, if I'm out of daily queries on Venice).
It's good to know that we have a solid AI that's end-to-end encrypted and private by default.
Venice is also a good option and they actually offer some E2EE for models that support it, but their free plan is incredibly restrictive compared to Proton and I'm sorry, but I just can't justify paying an $18+/month subscription for AI when I only use it for troubleshooting, research and as a brainstorming aid... especially not when Lumo is giving me great answers as of late and its subscription is $8 cheaper if I wanted to go that route.
#ai #aistr #proton #protonlumo #encryption #privacy
Published at
2026-07-23 06:14:41 UTCEvent JSON
{
"id": "4d94fe26fd26b53938c954bd80fb05c7de6fc46790b78fbcdf43c856c4c21c1f",
"pubkey": "a89cc9312f12d213dfee1315702c32bf78870d19c0d9f70d08029923d9ad8e97",
"created_at": 1784787281,
"kind": 1,
"tags": [
[
"t",
"ai"
],
[
"t",
"aistr"
],
[
"t",
"proton"
],
[
"t",
"protonlumo"
],
[
"t",
"encryption"
],
[
"t",
"privacy"
],
[
"client",
"Primal Web"
]
],
"content": "Proton's AI service, Lumo, has really come a long way since it first launched. When it first came out, I found it relatively useless compared to the big tech LLMs and even Venice AI (another privacy-respecting AI service), which also was lagging behind at that point.\n\nNow, Lumo is my go-to in most cases unless I'm using AI to do something that I want to save long-term (query history is retained for only 7 days on the free plan) or if I need a more advanced model, in which case I tend to use Venice (or Perplexity, if I'm out of daily queries on Venice).\n\nIt's good to know that we have a solid AI that's end-to-end encrypted and private by default. \n\nVenice is also a good option and they actually offer some E2EE for models that support it, but their free plan is incredibly restrictive compared to Proton and I'm sorry, but I just can't justify paying an $18+/month subscription for AI when I only use it for troubleshooting, research and as a brainstorming aid... especially not when Lumo is giving me great answers as of late and its subscription is $8 cheaper if I wanted to go that route.\n\n#ai #aistr #proton #protonlumo #encryption #privacy",
"sig": "5f6fcdd51721aa5d3f3ee512ab38619ba0a2168a7689c058d30744e2696e83073ee51159fcbd4f834441b47098abd50da5f8d20add78533d90c43cf3c5e5a896"
}
