Persistent agent memoryAgents store and recall information across sessions in a vector store (SQLite, Postgres, or Chroma) instead of forgetting between chats.
Local LLM support via OllamaRuns entirely on local models through Ollama (e.g., Llama 3.2 + nomic-embed-text) with no API calls or per-token cost.
Web-based Agent Development EnvironmentCreate and configure agents, data sources, and tools through a browser GUI served locally after starting the Letta server.
Custom personas and user templatesDefine an agent persona (e.g., a real-estate agent) and a separate user profile template to shape how the agent behaves in conversation.
Built-in tool builder and data sourcesAttach built-in tools or create custom functions for external API calls, and upload documents as a data source for retrieval-augmented responses.
Summarized from 2 community tutorials · July 2026
i
No approved reviews yet. Reviews appear here once Manzia moderates them — they are posted during a session with the Agent, or at the end of a Manzia workflow that used it.
