Top all-time · #5357
Letta
Company blog · letta.com/blogaiprogramming
Letta is the company blog at letta.com, writing about ai and programming. Since 2024, 15 different Hacker News readers have linked it 26 times, most often “Letta”. No feed was found, so read it on the site itself.
Too few sampled comments (8) to say how often HN recommends it.?
Why HN links it
HN readers link Letta, the open source agent framework formerly known as MemGPT, cited as a stateful memory and agent development tool, and as a source for research on agent memory and skill learning.
Summarised by Claude from a sample of 8 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Making machines that learn. Create stateful agents that remember everything, learn continuously, and improve themselves over time.
Most-linked posts?
- Letta6 people · 7 mentions · first linked 2025-01-14
- /blog/skill-learning2 people · 3 mentions · first linked 2025-12-16
What Hacker News says?
Letta | Multiple Roles | Onsite 5 days a week - San Francisco| Full-time | https://www.letta.com/ | $150k-250k + Equity Letta (fka MemGPT) is a open source agent development suite with three products, our open source agentic framework, our long-term stateful memory algorithm (and research) and a cloud and local agent development environment UI. We're popular with developers, hosting an active community of 1000+ on o…
Comment linking /
on “Ask HN: Who is hiring? (March 2025)”Submitted Letta Code as a story · 83 points
HN discussionMaybe this is a solution https://www.letta.com/
Comment linking /
on “LLM based agents as Dungeon Masters”Submitted Letta Agents SDK: An SDK for stateful agents as a story · 6 points
HN discussionI think the problem with ChatGPT / other RAG-based memory solutions is that it's not possible to collaborate with the agent on what it's memory should look like - so it makes sense that its much easier to just have a stateless system and message queue, to avoid mysterious pollution. But Letta's memory management is primarily text/files based so very transparent and controllable. An example of how this kind of memory…
Comment linking /blog/skill-learning
on “Letta Code”Submitted Context-Bench: Benchmarking LLMs on Agentic Context Engineering as a story · 5 points
HN discussionRegarding chess, I think that's an unfair challenge without access to long-term memory. Did you try to give it a memory scratchpad, e.g. with Letta? [0] https://www.letta.com/
Comment linking /
on “Gemini 2.5 gets 24.4% on MathArena USAMO beating previous top score of 4.7%”Submitted Continual Learning in Token Space as a story · 3 points
HN discussionThanks @dang for moderating! This is indeed not our original findings and this is a sub conclusion for an ablation we did to remove the confound of LLMs internal domain knowledge. Thanks for submitting for us @mustaphah here's a little bit more details on how we approach this: > I would frame the 'post-trajectory generated skills' as feedback-generated skills, so is Letta: https://www.letta.com/blog/skill-learning .…
Comment linking /blog/skill-learning
on “Study: Self-generated Agent Skills are useless”Submitted RAG is not agent memory as a story · 3 points
HN discussionI would frame the 'post-trajectory generated skills' as feedback-generated skills, so is Letta: https://www.letta.com/blog/skill-learning . We haven't seen existing research or hypothesis debating whether the skills improvement might come from the skill prompt themselves activated knowledge in LLMs that can help itself. So that's why we added an ablation of 'pre-trajectory generated skills' because we have that hypo…
Comment linking /blog/skill-learning
on “Study: Self-generated Agent Skills are useless”Submitted Continual Learning in Token Space (2025) as a story · 2 points
HN discussion> Hermes is one of these OpenClaw clones So that's what it is. Reading its README I thought it was another harness like Pi [1], but with built-in memory so it remembers what it learns, and gets more capable the longer it runs. Like Letta [2], Dirac [3][4] and the other "more experimental harnesses that look interesting but I haven't had time to try out". 1. https://pi.dev/ 2. https://www.letta.com/ 3. https://dirac.…
Comment linking /
on “HERMES.md: Anthropic bug causes $200 extra charge, refuses refund”Submitted The Letta Code App as a story · 2 points
HN discussionAn interesting idea, but I'm surprised that they only mention MemGPT/Letta [0] in passing; from what I've seen of Letta's work and e.g. their recent paper on "sleep-time compute", they are really pushing this space forward. [0] https://www.letta.com/ [1] https://arxiv.org/abs/2504.13171
Comment linking /
on “MemOS: A Memory OS for AI System”Submitted How to build the #1 open source terminal-use agent as a story · 2 points
HN discussionSubmitted Letta as a story · 2 points
HN discussionSubmitted Context Engineering for the LLM OS: User vs. Kernel Context as a story · 2 points
HN discussionSubmitted The software stack for LLM agents (in 2024) as a story · 2 points
HN discussionSubmitted Trajectory: A Standard Format for Agent Experience Data as a story · 1 points
HN discussionSubmitted Letta Agent – Personalized Agent That Remembers and Learns as a story · 1 points
HN discussionSubmitted Letta's Next Phase as a story · 1 points
HN discussionSubmitted Show HN: Letta – Git-Based Memory for Coding Agents as a story · 1 points
HN discussionSubmitted Context Repositories: Git-Based Memory for Coding Agents as a story · 1 points
HN discussionSubmitted Claude Sonnet 4.5 and the memory Omni-tool in Letta as a story · 1 points
HN discussionSubmitted Letta Leaderboard: Benchmarking LLMs on Agentic Memory as a story · 1 points
HN discussion
