For developers building AI agents, Letta is an agent runtime built around self-editing memory, giving agents state and context that persist and evolve across long-running interactions.
Presence & Market Position
Tracked · below panel floor — earned mentions are accruing toward the next scored window.
Measures earned, engagement-weighted share of voice across the GTM voices panel. Methodology →
Letta positions itself as an agent-runtime company built around self-editing memory, giving AI agents state and context that persist and evolve across long-running sessions instead of resetting each time. Its footprint in the broader GTM conversation is quieter so far, reflecting a category built around developer-facing agent infrastructure rather than go-to-market tooling itself. What visibility does exist centers on product and research milestones, including new model support, runtime capabilities, and integration updates, more than the comparison or workflow discussion that builds up around more established categories.
Deepline
Turbopuffer
Fluint
Cognee
ColdIQBilt, 11x, Hunt Club, Kognitos
Work at Letta? Claim this profile to add customers and case studies.
Business Profile
Letta came out of stealth in 2024 as a spinout from UC Berkeley's Sky Computing Lab, built around the MemGPT research on giving AI agents long-term memory. The company raised $10M in seed funding, has stayed independent since, and runs a team of 13. It has taken its self-improving agent runtime into production with a handful of early customers, including Bilt, 11x, Hunt Club, and Kognitos, and reported roughly $1.4M in ARR in 2025.
Agent Readiness
Measures how easily your agents can build on it — API, MCP, CLI, SDK, docs depth. Methodology →
Letta has built its runtime as multi-model and multi-surface from the outset: developers reach it through an API, SDK, and CLI, and recent updates added support for GPT- and Opus-generation models plus an OpenAI-compatible interface so stateful agents can be reached from third-party chat clients. A distributed-agent capability lets agents run on cloud sandboxes or remote servers with memory and context following them across machines, and a new extension system lets developers modify context and add custom tools.
IS THIS YOUR BRAND?
Claim this profile
Fact-check your data, add context, and earn the verified badge — free, takes minutes, work email required.
Claim Letta →