Versions:
AgentDB is a single-file embedded database developed by Datacules LLC and purpose-built for AI agents. The software combines several data management capabilities into one portable file, including SQL, Vector Search based on HNSW, Full-Text Search, Hybrid Queries, and Memory Graphs. According to the publisher's description, AgentDB operates with zero external dependencies, meaning it does not require additional software components or services to function. This design makes it suitable as an embedded storage and retrieval layer for applications centered on autonomous or semi-autonomous AI agents, where data persistence, retrieval, and relationship tracking must be handled locally within a single file. The current version of AgentDB is 0.7.0, and the catalog records three versions of the software in total, indicating ongoing development and iterative releases since its initial publication. The pre-1.0 version number reflects that the software is still in an early development stage, with functionality subject to refinement across its version history. In terms of categorization, AgentDB belongs to the database and data management category, more specifically the segment of embedded databases, and it also intersects with the artificial intelligence and machine learning tooling category due to its explicit orientation toward AI agent workloads. Its combination of SQL querying with vector similarity search positions it for use cases such as semantic retrieval, where embeddings must be searched efficiently using the HNSW algorithm, while Full-Text Search supports keyword-based lookup over textual content. Hybrid Queries allow these retrieval approaches to be combined within a single operation, and Memory Graphs provide a way to represent and traverse relationships between stored pieces of information, which is relevant to agent memory architectures. Because all of these capabilities are consolidated into one portable file with no external dependencies, AgentDB can be embedded directly into applications that need self-contained data storage and retrieval for AI-driven workflows. Organizations and developers evaluating embedded database options for AI agent projects may consider AgentDB among the tools addressing this specific combination of requirements.
Tags: