Types hub · 9 memory types
The Types of AI Agent Memory Explained
AI agent memory types include working (short-term), long-term, episodic, semantic and procedural memory — each storing different information and using different backends.
Memory type tiers
Why it matters
Why memory types matter for agent design
Different types require different storage, retrieval and update patterns. Map your data to the right type before choosing a framework.
Wrong type = wrong architecture
Episodic event logs need different retrieval than stable semantic facts. Procedural skills map to tools, not vector search alone.
Design by type first
Start with what the agent must remember, then pick backends and frameworks. → How AI memory works
Taxonomy
Main AI agent memory types
| Type | What it stores | Duration | Typical backend |
|---|---|---|---|
| Short-term / working | Current task, recent turns | Single session | Context window |
| Long-term | Cross-session knowledge | Persistent | Vector DB, graph, KV |
| Episodic | Specific past interactions | Persistent | Vector + metadata |
| Semantic | Stable facts and knowledge | Persistent | Vector DB, KG |
| Procedural | Skills and how-to procedures | Persistent | Tools, graph |
| Sensory / buffer | Raw inputs pre-processing | Milliseconds–seconds | In-memory buffer |
| Shared | Multi-agent common store | Persistent | Namespaced store |
Short-term memory
Working memory inside the context window.
Read guide →Long-term memory
Cross-session persistence — the foundation of stateful agents.
Read guide →Episodic memory
Specific past interactions the agent can reference.
Read guide →Semantic memory
Stable facts and learned knowledge.
Read guide →Procedural memory
Skills, tools and learned procedures.
Read guide →Sensory & buffer memory
Brief holding area for raw inputs.
Read guide →Parametric vs non-parametric
Weights vs external store — the key distinction.
Read guide →Short-term vs long-term
The two-tier model and promotion via consolidation.
Read guide →Shared memory
Common store for multi-agent systems.
Read guide →Disambiguation
Human memory types vs AI agent memory
This taxonomy is inspired by cognitive science — but AI agent memory is implemented in software, not biology. It is not computer hardware RAM or LSTM networks.
| Human / cognitive term | AI agent equivalent |
|---|---|
| Working memory | Context window + recent turns |
| Long-term memory | External vector/graph/KV store |
| Episodic memory | Interaction logs with timestamps |
| Semantic memory | Embedded facts and preferences |
| Procedural memory | Tool skills and learned workflows |
FAQ
Frequently asked questions
How many types of AI agent memory are there?
The core taxonomy has six types: short-term/working, long-term, episodic, semantic, procedural and sensory/buffer — plus cross-cutting distinctions like parametric vs non-parametric and shared memory for multi-agent systems. See the types hub.
What is the difference between short-term and long-term memory in AI agents?
Short-term (working) memory lives in the context window and holds the current task. Long-term memory persists in an external store across sessions. Agents promote information between tiers via consolidation. See short-term vs long-term.
Episodic vs semantic memory — what's the difference?
Episodic memory stores specific past interactions ("user asked about a refund on Tuesday"). Semantic memory stores stable facts and preferences ("user prefers email"). Both persist long-term but serve different retrieval patterns.
What is working memory in AI agents?
Working memory is the agent's short-term store — typically the context window holding recent turns, tool outputs and the current task. It is lost when the session ends unless written to long-term memory. See short-term memory.
What is procedural memory in AI agents?
Procedural memory stores skills, tools and learned how-to procedures — not facts but capabilities. It bridges to memory as a tool patterns. See procedural memory.
Parametric vs non-parametric memory — which do agents use?
Most agent memory is non-parametric — stored externally and updatable at runtime. Parametric memory lives in model weights (fine-tuning). See parametric vs non-parametric.
Is LSTM the same as AI agent memory?
No. LSTM is a neural network architecture for sequence modeling — not the cognitive-science-inspired memory taxonomy used for AI agents. Agent memory refers to external storage systems (vector DBs, graphs, APIs), not recurrent network cells.
Do human memory types apply to AI agents?
The taxonomy is inspired by cognitive science but implemented differently in software. Human memory is biological; agent memory uses context windows, vector stores and knowledge graphs. See AI memory vs human memory.
Which memory type is best for personalization?
Semantic memory (stable user preferences and facts) plus episodic memory (past interactions) power personalization. Frameworks like Engram and Mem0 specialize in per-user semantic memory. See user memory personalization.
How do memory types map to Mem0, Zep and Letta?
Engram maps semantic/episodic memory on Weaviate-native stacks. Mem0 focuses on per-user semantic/episodic vector memory. Zep adds temporal knowledge-graph semantics. Letta uses virtual-context paging across tiers. See best AI memory tools for the full mapping.
Continue exploring
Architecture, frameworks and implementation guides.