--- title: "Agent Memory you can trust | SurrealDB" description: "Durable memory for AI agents: recall with provenance on every fact, built on SurrealDB. Join the waitlist for early access; plans from $29/month at launch." url: https://surrealdb.com/agent-memory --- # Agent Memory you can trust AI agents forget everything when a session ends. Agent Memory is the memory layer that gives them durable, long-term memory and recall you can trust - so every agent you build remembers what happened, and can prove where each answer came from. [Join the waitlist](https://surrealdb.com/agent-memory/get-started) [Book a demo](https://surrealdb.com/contact/sales) Works with · [+29 more](https://surrealdb.com/agent-memory?__original_path=%2Fagent-memory.md#integrations) ![Claude Desktop and Claude Code](https://surrealdb.com/assets/static/claude.qeceRvhT.svg)![Cursor](https://surrealdb.com/assets/static/cursor.CffcCNKv.svg)![OpenAI Codex](https://surrealdb.com/assets/static/openai.CNSrSktx.svg)![OpenClaw](https://surrealdb.com/assets/static/openclaw.CzSOJJfQ.svg)![Hermes](https://surrealdb.com/assets/static/hermes.u8A4-1Lq.svg)![VS Code](https://surrealdb.com/assets/static/vscode.ofeWxTsy.svg)![Windsurf](https://surrealdb.com/assets/static/windsurf.Ww98i45f.svg)![Zed](https://surrealdb.com/assets/static/zed.M8dgM-za.svg)![LangChain](https://surrealdb.com/assets/static/langchain.B2qiXaj0.svg)![Vercel AI SDK](https://surrealdb.com/assets/static/vercel.4jR_VMI9.svg)![Cloudflare](https://surrealdb.com/assets/static/cloudflare.CGYztXwG.svg)![n8n](https://surrealdb.com/assets/static/n8n.CjHLK_Mt.svg)![Zapier](https://surrealdb.com/assets/static/zapier.V_eYzQMe.svg) ``` const memory = new AgentMemory({ context: "acme-prod",});await memory.remember( "I work at Acme as CTO",);const hits = await memory.recall( "what do I do at Acme",); ``` hits Works at Acme as CTO Session 4 · turn 12 · confidence 0.94 Prefers squash merges on this repo Session 9 · turn 3 · confidence 0.88 1. ![Babcock](https://surrealdb.com/assets/static/babcock.lo4rnVg1.svg) 2. ![ING](https://surrealdb.com/assets/static/ing.X3I6S3_V.svg) 3. ![British Airways](https://surrealdb.com/assets/static/british-airways.KEsZiwV-.svg) 4. ![Nvidia](https://surrealdb.com/assets/static/nvidia.DaIEuMil.svg) 5. ![Apple](https://surrealdb.com/assets/static/apple.D5pq4flY.svg) 6. ![SpaceX](https://surrealdb.com/assets/static/spacex.CQJEk-IL.svg) 7. ![Samsung](https://surrealdb.com/assets/static/samsung.CH-vQgnb.svg) 8. ![adidas](https://surrealdb.com/assets/static/adidas.DdTC5qhk.svg) 9. ![Tencent](https://surrealdb.com/assets/static/tencent.paQmLxyy.svg) 10. ![Alibaba](https://surrealdb.com/assets/static/alibaba.B16idgfM.svg) 11. ![PolyAI](https://surrealdb.com/assets/static/poly-ai.c3w_fAg6.svg) 12. ![Later](https://surrealdb.com/assets/static/later.Ds736jFO.svg) 13. ![Verizon](https://surrealdb.com/assets/static/verizon.BI7CajdX.svg) 14. ![Liberty Mutual](https://surrealdb.com/assets/static/liberty-mutual.B7qOU1pd.svg) 15. ![Walmart](https://surrealdb.com/assets/static/walmart.BjDg_Sr8.svg) 16. ![Carrier](https://surrealdb.com/assets/static/carrier.D21gC6NX.svg) 17. ![Saks Fifth Avenue](https://surrealdb.com/assets/static/saks-fifth-avenue.COIDpLSb.svg) 18. ![San Francisco Compute Company](https://surrealdb.com/assets/static/sfcc.B7jlImq4.svg) 19. ![Shield AI](https://surrealdb.com/assets/static/shield-ai.pINZ0KJr.svg) 20. ![Wix](https://surrealdb.com/assets/static/wix.DvHhmoBi.svg) WHY AGENTS FORGET ## Agents fail because of context, not models. An agent's only built-in memory is the context window, and it is wiped when the session ends. Everything it worked out, it works out again - the same files re-read, the same dead ends re-tried, the same corrections re-typed by whoever is supervising it. Session 1 · what it learned Token-refresh race Squash-merges, tabs Cache rewrite: ruled out Session ends context window cleared Without memory Starts over Re-reads every file Re-tries the discarded approach Asks the same questions again With Agent Memory Picks up instantly Token-refresh race Session 1 · turn 8 Squash-merges, tabs Session 1 · turn 14 Cache rewrite: ruled out Session 1 · turn 31 Same model, same prompt, same tools. The only difference is whether the last session survived - and, because every recalled fact cites the turn it came from, whether the agent can show you why it believes it. THE SUITE ## One memory layer, six moving parts. Memory, a knowledge graph, hybrid recall, tiered queries, provenance, and background understanding - not six products to integrate, but one system you buy, secure, and audit once. ### [Typed memory](https://surrealdb.com/agent-memory/deep-dive#memory-model) Six categories extracted from every turn - episodic, identity, knowledge, context, instructions, and uncertainty - each with its own lifecycle and retrieval weight. Not one undifferentiated blob. Memory model ### [Knowledge graph](https://surrealdb.com/agent-memory/deep-dive#document-pipeline) Entities as nodes, typed relationships as edges. Because edges are documents, they carry their own properties, confidence, and timestamps - and uploaded documents become structure, not a blob in a vector store. Document pipeline ### [Hybrid recall](https://surrealdb.com/agent-memory/deep-dive#hybrid-retrieval) Eight signals fused in one auditable ranker - vector, BM25, graph traversal, keyword bridges, document links, PageRank, geographic recall, and trace-derived features. Not similarity alone. Hybrid retrieval ### [Tiered queries](https://surrealdb.com/agent-memory/deep-dive#tiered-queries) Cheap questions stay cheap. A typed lookup answers straight from the graph and never touches the ranker; only real retrieval runs the full fusion, with a broader sweep held in reserve. Query tiers ### [Provenance and time](https://surrealdb.com/agent-memory/deep-dive#audit-time) Every fact knows where it came from and when it was true. Agent Memory supersedes instead of deletes: the old fact stays, the new one wins, and every answer points back to the exact messages behind it. Audit and time ### [Autonomous understanding](https://surrealdb.com/agent-memory/deep-dive#autonomous-understanding) Memory that improves between conversations. Reflection, elaboration, and consolidation run in the background, so the graph gets sharper without anyone prompting it to. Background work HOW IT WORKS ## Every turn: read → think → write Agents operate in a continuous cycle: read context, reason over it, and write results back. Agent Memory keeps the entire loop inside one engine and one transaction boundary. Read SurrealQL Multi-model query Graph Vector Temporal Unified result set Pull the whole context in one round trip Graph traversal, vector search, and temporal facts resolve in a single SurrealQL statement. No fan-out across a vector store, a graph store, and a row store, and no application-side join to reconcile them. One statement, one round trip Eight retrieval signals fused in one ranker Per-verb grants applied at the engine, deny by default Think Result set Multi-model query output Documents Relations Vectors History Complete context The agent reasons over complete context Documents, relationships, embeddings, and history arrive together rather than as separate results the agent has to stitch. What comes back is the slice that matters for this turn, with the sources attached. Documents, relations, and embeddings in one payload Superseded facts marked, not silently dropped Every answer carries the exact sources behind it Write BEGIN TRANSACTION ACID transaction boundary Update Create Trigger COMMITTED Persist the turn in one ACID transaction Decisions, entity updates, and events commit together or not at all. There is no window where the vector index has the embedding but the graph has not yet learned the fact. One ACID transaction per fact No partial writes, no cross-store drift Tri-temporal history recorded on every change FOR DEVELOPERS ## Drop it in beneath your agent. Agent Memory doesn't replace your agent framework or manage the context window - it's the memory layer you add underneath. Connect over MCP from any client, the TypeScript SDK, the CLI, or the REST API, with adapters for LangChain, OpenAI Agents, the Vercel AI SDK, and n8n. Then it's two calls: remember and recall. Applications ![Agent Memory](https://surrealdb.com/assets/static/surrealdb-icon.C3ORaDk9.svg) Agent Memory Entity extraction Knowledge graph Temporal facts Hybrid retrieval ![SurrealDB](https://surrealdb.com/assets/static/surrealdb-icon.C3ORaDk9.svg) Database Documents Graphs Vectors Time-series Auth APIs Distributed write nodes Node A Node B Node C Object storage (S3 / S3-compatible) TypeScript SDK MCP CLI REST ``` // npm install @surrealdb/agent-memoryimport { AgentMemory } from "@surrealdb/agent-memory";const memory = new AgentMemory({ context: "acme-prod" });await memory.remember("I work at Acme as CTO");const hits = await memory.recall("what do I do at Acme"); ``` [Read the docs](https://surrealdb.com/docs/agent-memory) [See every integration](https://surrealdb.com/agent-memory?__original_path=%2Fagent-memory.md#integrations) One Rust binary with MCP built in - seven tools: remember, recall, context, reflect, forget, upload, and inspect. No Python runtime, no sidecars. ONE SUBSTRATE ## One engine, one transaction. Most memory layers stitch a vector store, a graph store, and a row store together with glue code, and inherit every seam - no cross-store transactions, drift, separate scaling. Agent Memory is one SurrealDB engine: graph, vector, document, and time in a single ACID transaction, no plumbing to run. Before - Multiple stores Client 0ms Your application 6 SDKs · 6 drivers · 6 connection pools Vector database Proprietary API Full-text search database Search DSL Time-series database InfluxQL / Flux Document database SQL / MQL Graph database Cypher / Gremlin Relational database SQL Eventual consistency only After - Agent Memory on SurrealDB Client 0ms Your application 1 SDK · 1 driver · 1 connection pool ![SurrealDB](https://surrealdb.com/assets/static/surrealdb-icon.C3ORaDk9.svg) Database SurrealQL Vectors Full-text Time-series Documents Graphs Relational Single ACID transactionStrong consistency ### The architecture, row by row Agent Memory Memory middleware Vector databases Storage model One multi-model database Two or three stitched stores Vector index only Write consistency One ACID transaction per fact Eventual across stores Not included Provenance Stored field on every row Bolt-on metadata Not included Time model Tri-temporal: system, known, valid Timestamps Not included Audit trail Queryable trace graph External logging Not included Retrieval signals Eight, fused in one ranker Vector plus rerank Similarity only Access control Per-verb grants, deny by default Application-level Application-level Extra infrastructure None - scales to zero A service plus its stores Plus a database for state PRICING ## Simple plans, priced on usage. Three self-serve plans priced on the tokens your agents actually use, and an enterprise package for larger allowances and single-tenant isolation. SurrealDB, the engine underneath, is open source and free to self-host. One-week free trial Full memory layer, no feature gates MCP, SDK, CLI, and REST Scales to zero between sessions Agent Memory Lite ## $29 /mo The entry rung of the ladder - the full memory layer, sized for getting started. What you get 1 million tokens per month included One-week free trial Knowledge graph, provenance, and tri-temporal memory Hybrid retrieval (vector + BM25 + graph) MCP server, TypeScript SDK, and CLI Plus Community support [Join the waitlist](https://surrealdb.com/agent-memory/get-started) Agent Memory ## $299 /mo For production agent workloads with real usage. What you get 10 million tokens per month included One-week free trial Everything in Agent Memory Lite Plus Community support [Join the waitlist](https://surrealdb.com/agent-memory/get-started) Agent Memory+ ## $1,099 /mo For scaling agent fleets and heavier ingestion. What you get 40 million tokens per month included One-week free trial Everything in Agent Memory Plus Community support [Join the waitlist](https://surrealdb.com/agent-memory/get-started) Enterprise ## Custom For larger token needs and enterprise controls. What you get Larger token allowances Single-tenant on SurrealDB Cloud Self-hosted and air-gapped deployments for qualifying engagements Custom SLAs Plus Enterprise support [Talk to sales](https://surrealdb.com/contact/sales) Usage beyond the included tokens is charged at the published overage rate for your plan. Enterprise allowances, single-tenant deployments, and custom SLAs are agreed with sales. ### Plans and usage Agent Memory Lite Agent Memory Agent Memory+ Enterprise Monthly price $29 $299 $1,099 Custom Included tokens per month 1 million 10 million 40 million Custom One-week free trial Included Included Included Not included Usage beyond included tokens Overage rates apply Overage rates apply Overage rates apply Custom ### Memory capabilities Agent Memory Lite Agent Memory Agent Memory+ Enterprise Knowledge graph with entity extraction Included Included Included Included Conversational memory Included Included Included Included Tri-temporal versioning and supersession Included Included Included Included Provenance on every fact Included Included Included Included Hybrid retrieval (vector + BM25 + graph) Included Included Included Included Retrieval, decision, and response traces Included Included Included Included Multi-modal document pipeline Included Included Included Included MCP server, TypeScript SDK, and CLI Included Included Included Included ### Isolation and security Agent Memory Lite Agent Memory Agent Memory+ Enterprise Context isolation (own namespace and database) Included Included Included Included Scoped permissions and API-key attenuation Included Included Included Included Prompt-injection scanning Included Included Included Included First-class forget verb (GDPR erasure) Included Included Included Included Single-tenant deployment Not included Not included Not included Included Self-hosted / air-gapped deployment Not included Not included Not included Qualifying engagements ### Support Agent Memory Lite Agent Memory Agent Memory+ Enterprise Community support Included Included Included Included Enterprise support with SLAs Not included Not included Not included Included INTEGRATIONS ## Connects to everything you already use MCP clients, the TypeScript SDK, the CLI and REST API, agent frameworks and automation platforms - the same wall as the product's own connect surface. SDKs & API Talk to a Context directly from your own code. Agent Memory CLI Connect via CLI JavaScript Connect via SDK Python Connect via SDK Go Connect via SDK Swift Connect via SDK Kotlin Connect via SDK Haskell Connect via SDK Elixir Connect via SDK Dart Connect via SDK REST API Connect via API Agents & coding tools Connect your AI agents and coding assistants for persistent memory. Claude Code Connect via MCP Claude Desktop Connect via MCP Cursor Connect via MCP OpenAI Codex Connect via MCP OpenClaw Connect via MCP Hermes Connect via MCP VS Code Connect via MCP Windsurf Connect via MCP Zed Connect via MCP Any MCP client Connect via MCP Frameworks Wire a Context into your agent framework of choice. LangChain Connect via package OpenAI Agents Connect via package Vercel AI SDK Connect via package EveJS Connect via package Cloudflare Connect via MCP TanStack AI Connect via MCP Mastra Connect via MCP Automation platforms Give your automation workflows access to a Context's memory. n8n Connect via node Zapier Connect via MCP Messaging Bring a Context's memory to the channels you already message on. WhatsApp Coming soon Telegram Coming soon Facebook Messenger Coming soon SMS / RCS Coming soon Connectors Sync memory to and from the tools and data platforms your team uses. SurrealDB Coming soon Slack Coming soon Notion Coming soon Google Drive Coming soon GitHub Coming soon Linear Coming soon Confluence Coming soon Databricks Coming soon Snowflake Coming soon GET STARTED ## Give your agents a memory that lasts Durable memory, trustworthy recall, and a knowledge graph - one system you buy, secure, and audit once. Built on SurrealDB. ![Samsung](https://surrealdb.com/assets/static/4c58b81e7b3c9466.C_Hv0eml.svg)![NVIDIA](https://surrealdb.com/assets/static/nvidia.DaIEuMil.svg)![Apple](https://surrealdb.com/assets/static/f7dc2519e0d212bc.Cn8MYAK7.svg)![Verizon](https://surrealdb.com/assets/static/18b99996c689000f.B5PQ-nI9.svg)![Tencent](https://surrealdb.com/assets/static/401d8346058682c8.DqM87mst.svg) SOC 2 Type 2 GDPR Cyber Essentials Plus ISO 27001 At launch, every plan starts with a one-week free trial · self-serve from $29 a month. [Join the waitlist](https://surrealdb.com/agent-memory/get-started) [Talk to sales](https://surrealdb.com/contact/sales) FREQUENTLY ASKED QUESTIONS ## Agent Memory FAQ How do I get started with Agent Memory? How is Agent Memory different from Mem0, Zep, or Letta? How is Agent Memory different from a vector database like Pinecone or Weaviate? How is Agent Memory different from specialist agent databases? Why not just use Postgres with pgvector? Does Agent Memory replace my agent runtime or manage the context window? How does Agent Memory prove that an answer is grounded in the substrate? Is Agent Memory open source? How much does Agent Memory cost? Where does my data go, and can I delete it on request? ```json {"@context":"https://schema.org","@type":"Organization","name":"SurrealDB","url":"https://surrealdb.com","logo":"https://surrealdb.com/assets/static/logo.BG7_TG2b.svg","description":"SurrealDB is the unified data layer for AI. A multi-model database for documents, graphs, vectors, and time-series.","foundingDate":"2022","hasCertification":[{"@type":"Certification","name":"SOC 2 Type 2"},{"@type":"Certification","name":"GDPR"},{"@type":"Certification","name":"Cyber Essentials Plus"},{"@type":"Certification","name":"ISO 27001"}],"owns":[{"@type":"SoftwareApplication","name":"SurrealDB","url":"https://surrealdb.com/surrealdb"},{"@type":"SoftwareApplication","name":"Agent Memory","url":"https://surrealdb.com/agent-memory"}],"knowsAbout":["multi-model databases","document databases","graph databases","vector search","time-series databases","SurrealQL","Agent Memory","real-time databases","embedded databases","context layer","graph ontology","distributed database","knowledge graphs","distributed transaction protocols","highly-scalable databases"],"sameAs":["https://www.wikidata.org/wiki/Q124316308","https://github.com/surrealdb/surrealdb","https://twitter.com/surrealdb","https://www.youtube.com/@surrealdb","https://www.linkedin.com/company/surrealdb","https://discord.gg/surrealdb","https://www.reddit.com/r/surrealdb","https://www.instagram.com/surrealdb","https://medium.com/surrealdb","https://dev.to/surrealdb"]} ``` ```json {"@context":"https://schema.org","@type":"SoftwareApplication","name":"Agent Memory","applicationCategory":"DeveloperApplication","operatingSystem":"Cloud","url":"https://surrealdb.com/agent-memory","description":"Agent Memory is a memory layer for AI agents: it remembers what happened, tracks when the facts changed, and can prove where every answer came from. Join the waitlist for early access; plans from $29 a month at launch. Built on SurrealDB.","publisher":{"@type":"Organization","name":"SurrealDB","url":"https://surrealdb.com"},"offers":[{"@type":"Offer","name":"Agent Memory Lite","availability":"https://schema.org/PreOrder","url":"https://surrealdb.com/agent-memory/get-started","priceSpecification":{"@type":"UnitPriceSpecification","price":"29","priceCurrency":"USD","referenceQuantity":{"@type":"QuantitativeValue","value":1,"unitCode":"MON"}}},{"@type":"Offer","name":"Agent Memory","availability":"https://schema.org/PreOrder","url":"https://surrealdb.com/agent-memory/get-started","priceSpecification":{"@type":"UnitPriceSpecification","price":"299","priceCurrency":"USD","referenceQuantity":{"@type":"QuantitativeValue","value":1,"unitCode":"MON"}}},{"@type":"Offer","name":"Agent Memory+","availability":"https://schema.org/PreOrder","url":"https://surrealdb.com/agent-memory/get-started","priceSpecification":{"@type":"UnitPriceSpecification","price":"1099","priceCurrency":"USD","referenceQuantity":{"@type":"QuantitativeValue","value":1,"unitCode":"MON"}}}],"isBasedOn":"https://surrealdb.com/surrealdb"} ``` ```json {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://surrealdb.com"},{"@type":"ListItem","position":2,"name":"Agent Memory","item":"https://surrealdb.com/agent-memory"}]} ``` ```json {"@context":"https://schema.org","@type":"FAQPage","mainEntity":[{"@type":"Question","name":"How do I get started with Agent Memory?","acceptedAnswer":{"@type":"Answer","text":"Join the waitlist for early access. At launch, every self-serve plan begins with one free week. You create an Agent Memory Context on SurrealDB Cloud and connect: from an MCP client (Cursor, Claude Desktop, Claude Code) it is a single configuration entry; from application code, install the TypeScript SDK (@surrealdb/agent-memory); from the terminal, install the Agent Memory CLI with one command. Harness adapters for LangChain, OpenAI Agents, Vercel AI SDK, and n8n flush conversations into the substrate automatically, and Zapier connects via MCP."}},{"@type":"Question","name":"How is Agent Memory different from Mem0, Zep, or Letta?","acceptedAnswer":{"@type":"Answer","text":"Memory middleware libraries orchestrate two or more separate stores - typically a vector database, a graph database, and a row store - and absorb the consistency gaps in application code. Agent Memory is the database: graph, vector, document and structured records commit in one ACID transaction. Authoritative knowledge (uploaded documents) and experiential memory (conversational input) flow through the same reconciler into a single entity and relation graph, distinguished by provenance rather than by a separate store. Multi-tenancy, supersession, tri-temporal history, and trace-graph audit are engine-enforced, not application code."}},{"@type":"Question","name":"How is Agent Memory different from a vector database like Pinecone or Weaviate?","acceptedAnswer":{"@type":"Answer","text":"Agent Memory is not a vector database. Vectors are one signal among eight in the fused ranker - alongside BM25, graph traversal, keyword bridges, document links, PageRank, geographic recall, and trace-derived features. Embeddings-only retrieval has known failure modes (near-duplicates dominate top-k, rare-term queries miss the right chunk, structure between facts is invisible). Agent Memory's structural index is built at write time and read cheaply on every retrieval, with per-feature scores recorded on every trace - so retrieval is auditable, not black-box top-k."}},{"@type":"Question","name":"How is Agent Memory different from specialist agent databases?","acceptedAnswer":{"@type":"Answer","text":"Specialist agent databases saw, correctly, that flat vector stores fail for agent memory. But they respond by internalising a multi-store architecture instead of eliminating it: a vector index, a graph store, and a row store coordinated inside one product, with the seams moved rather than removed. Agent Memory is one engine - graph, vector, document and structured records commit in one ACID transaction, under one permission model, with provenance and tri-temporal history on every fact."}},{"@type":"Question","name":"Why not just use Postgres with pgvector?","acceptedAnswer":{"@type":"Answer","text":"pgvector adds vector similarity to Postgres, but agent memory needs more than similarity. There is no native graph to traverse, no tri-temporal versioning to answer 'what did we believe on date X', and no unified transaction across data models - so the glue code between extensions and sidecar stores becomes the product you maintain. Agent Memory is built on SurrealDB, where graph, vector, document and structured records are one engine and one transaction, with supersession, provenance, and audit built in."}},{"@type":"Question","name":"Does Agent Memory replace my agent runtime or manage the context window?","acceptedAnswer":{"@type":"Answer","text":"No. Agent Memory is the durable memory and knowledge layer behind an agent, not the in-prompt scratchpad and not the orchestration loop. The agent decides what to put in the context window; Agent Memory decides what to remember. POST /chat is a convenience composition over /facts and /query for callers who want Agent Memory to be an agent, but model orchestration, tool calling, and planning loops stay with your runtime of choice - including LangChain, OpenAI Agents, Vercel AI SDK, and n8n; MCP clients such as Claude Code connect directly."}},{"@type":"Question","name":"How does Agent Memory prove that an answer is grounded in the substrate?","acceptedAnswer":{"@type":"Answer","text":"Every fact-bearing row carries a source object: kind, ref, trust, lexical span (byte offsets into the originating turn or document), location, and derived_from. Every read produces a retrieval_trace listing the candidates considered and the rows returned; every decision produces a decision_trace recording created, superseded, and flagged rows; every response produces a response_trace linked to the retrievals and decisions behind it. Together they form a graph-resident audit log - any answer can be walked back to the retrievals it used, the entities those retrievals returned, and the bytes those entities cited."}},{"@type":"Question","name":"Is Agent Memory open source?","acceptedAnswer":{"@type":"Answer","text":"SurrealDB, the database engine underneath, is open source and free to self-host. The Agent Memory layer on top is closed source and ships as a single Rust binary with no Python runtime and no sidecars. Our roadmap intent is to upstream foundational parts of the Agent Memory model into SurrealDB over time so the most fundamental primitives remain open."}},{"@type":"Question","name":"How much does Agent Memory cost?","acceptedAnswer":{"@type":"Answer","text":"There are three self-serve plans, priced on the tokens your agents use: Agent Memory Lite at $29 a month for 1 million tokens, Agent Memory at $299 a month for 10 million, and Agent Memory+ at $1,099 a month for 40 million. Each starts with a one-week free trial, and each includes the full memory layer with no feature gates - the tiers differ on volume, not capability. Usage beyond the included tokens is charged at the published overage rate for your plan. Enterprise pricing covers larger allowances, single-tenant deployments on SurrealDB Cloud, self-hosted and air-gapped options for qualifying engagements, and custom SLAs. SurrealDB, the database engine underneath, is open source and free to self-host."}},{"@type":"Question","name":"Where does my data go, and can I delete it on request?","acceptedAnswer":{"@type":"Answer","text":"Agent Memory runs on SurrealDB Cloud today, with enterprise deployments running single-tenant. Tenant isolation is at the substrate level: each Context is its own SurrealDB namespace and database, and cross-Context reads are not expressible in the API. The four LLM hooks (extraction, reconciliation, synthesis, background) are configurable per Context, including any OpenAI-compatible, Anthropic, Google, or local provider. The embedding model is fixed to Google gemini-embedding-2 at launch, so embedding calls Google's API; per-Context embedding selection is on the roadmap. Self-hosted and air-gapped deployments, with local model inference, are available for qualifying enterprise engagements. forget is a first-class verb, distinct from natural aging: it removes derived rows (entities, attributes, relations, and any traces touching the target), the originating turn or document if requested, and the object-store original by content hash. Supersession history is retained by default for audit; a --purge flag removes that too."}}]} ```