StoryFirst seen
Devin builds a memory graph of user preferences across sessions and prunes stale records overnight via an open Agent Memory Repo standard.
Everything it touches
4 names 3 voices 3 source posts
The story’s neighborhood
Explore the graph, or choose a connection below.
Follow the evidence
This brief comes from Tech Twitter’s edited headlines. The posts below are the sources grouped with the story; statements in them belong to their authors.
Introducing Dreaming: Across sessions, Devin builds a memory graph of how you like to work At night Devin self-improves its memory to remove stale records and discover latent information We are creating an OSS standard called Agent Memory Repo:Read original on X
cool to see cognition doing "dreaming" - memory needs an offline cleanup loop, not just better retrieval, otherwise stale records pile up open q is how inferred memories get validated before the agent uses them glad it's an open standard too - memory should be openRead original on X
Today we shipped Memory and Dreaming in Devin: persistent self-cleaning memory for agent sessions. The standard is OSS and also works across any harness in any environment. ⚡Read original on X
Cognition ships Dreaming, self-cleaning session memory for Devin
Cognition ships Dreaming, self-cleaning session memory for Devin
Cognition ships Dreaming, self-cleaning session memory for Devin
Cognition ships Dreaming, self-cleaning memory for Devin agents
Cognition ships Dreaming for Devin's overnight agent memory cleanup
Cognition ships Dreaming for Devin's overnight agent memory cleanup
Cognition launches Devin Dreaming for overnight agent memory cleanup
Cognition launches Devin Dreaming for overnight agent memory cleanup
Cognition launches Devin Dreaming for overnight agent memory cleanup
Cognition launches Devin Dreaming for overnight agent memory cleanup
Cognition launches Devin Dreaming for overnight agent memory cleanup
Cognition launches Devin Dreaming for persistent agent memory
Cognition launches Devin Dreaming for persistent agent memory
Shared connections
The model is tuned for real-world coding tasks, giving developers another option among AI coding tools.
The post contrasts Devin's per-channel agents with a proposed single router, showing continuous agents are already live.
It ranks near the top at $4.04 per task, behind three rival models in the same evaluation.
OpenAI is pushing cheaper models to win these large accounts, which makes Anthropic's lead harder to challenge.