TencentDB Agent Memory is a team-level memory infrastructure for AI agents that automatically extracts and manages long-term memory assets from workflows.
Follow TencentDB Agent Memory to get an email when its listing changes, such as new pricing or features. One digest a day at most, and you can unsubscribe anytime.
It provides a four-layer architecture (L0-L3) to store and recall structured information, ensuring knowledge continuity across sessions, tasks, and teams.
- 01
Four-layer progressive memory architecture (L0-L3) for structured storage and recall
- 02
Dual-path retrieval mechanism combining Embedding and keyword search
- 03
Automatic extraction of four memory asset types: Chat Memory, Skill, Wiki, and CodeGraph
- 04
Memory Hub for centralized asset governance, ACL management, and team collaboration
- 05
Transparent Memory Proxy compatible with OpenAI and Anthropic protocols
Paid
Billed hourly based on actual model invocation compute (Credit) and memory data storage usage. Specific rates are 0.02 CNY per 10,000 items/hour for storage and 0.01 CNY per Credit/hour for model calls.
Best for
Development teams and enterprises needing to share project context, code knowledge, and operational expertise across multiple AI agents to reduce repetitive work and improve task completion rates.
Not ideal for
Simple, short-lived, or isolated tasks that do not require knowledge persistence, cross-session context, or collaborative team-based asset management.
An advanced, enterprise-grade memory orchestration layer that effectively solves the cold-start problem for AI agents by turning historical work traces into reusable, governed assets.
Version 2.0.3-beta.1 (2026-10-13) introduced a pluggable decision service and session trajectory recording. Version 2.0.2 (2026-10-09) added experimental MongoDB backend support and additional client integrations. Version 2.0.0 (2026-08-03) marked the official release featuring full open-source support for the four memory asset types and the Memory Hub management panel.
As of
Is TencentDB Agent Memory free?+
TencentDB Agent Memory is Paid. Billed hourly based on actual model invocation compute (Credit) and memory data storage usage. Specific rates are 0.02 CNY per 10,000 items/hour for storage and 0.01 CNY per Credit/hour for model calls. Check the official site for current plans.
Is TencentDB Agent Memory open source?+
TencentDB Agent Memory is not listed as open source on ai.dosa.dev - it is Paid.
How much does TencentDB Agent Memory cost?+
Billed hourly based on actual model invocation compute (Credit) and memory data storage usage. Specific rates are 0.02 CNY per 10,000 items/hour for storage and 0.01 CNY per Credit/hour for model calls. Pricing changes often - verify on the official site.
Who is TencentDB Agent Memory best for?+
Development teams and enterprises needing to share project context, code knowledge, and operational expertise across multiple AI agents to reduce repetitive work and improve task completion rates.
Who is TencentDB Agent Memory not ideal for?+
Simple, short-lived, or isolated tasks that do not require knowledge persistence, cross-session context, or collaborative team-based asset management.
What are the best TencentDB Agent Memory alternatives?+
The closest TencentDB Agent Memory alternatives on ai.dosa.dev are Pieces for Developers, agenttrace, Vibe Kanban - all listed under Developer Productivity & Workflow.
Who makes TencentDB Agent Memory?+
TencentDB Agent Memory is developed by Tencent Cloud. It is listed in the Developer Productivity & Workflow category on ai.dosa.dev.
Favorite this tool to revisit it later, or Zap it to contribute to the public vote count.