Mastra Memory
https://mastra.ai/docs/memory/overview
TypeScript-first agent framework. Default memory: LibSQL store + on-device fastembed-js for semantic recall.
Active · Verified 2026-10-22 · fresh as of 2026-10-22 · 66 cited facts, 3 with a third-party source · 1 recorded change · Suggest a correction
At a glance
- Type
- LibSQL + semantic recall vendor fix
- Tier
- T2
- Section
- Framework-embedded memory
- Latest release
- @mastra/core v1.79.0 (2026-10-21) repo fix
- Pricing
- Free + paid vendor fix
- Funding
- $35M total Series A · 2026-01 vendor fix
- Freshness
- 2026-10-22 fix
- Layer
- agent-infra repo fix
- SDK languages
- typescript repo fix
- Protocols
- mcp-client, mcp-server vendor fix
- Developer activity
- OpenRank 118.19 (12-mo avg 109.68) · 403 participants — 2026-09 third-party fix
Memory taxonomy
- storage
- relational
- retrieval
- similarity
- persistence
- cross-session
- update
- append-only
- unit
- episode
- governance
- inspectable
- conflict
- none
When to use
Optimised for: TypeScript-native developer experience + LibSQL semantic recall
Anti-fit: not for Python-only stacks
Pros & cons
Pros
Modern TypeScript-first framework with memory built in; LibSQL default substrate keeps ops simple.
Cons
TypeScript ecosystem — less appealing for Python-only shops.
Claims & capabilities
"Observational Memory" reports 94.87% on LongMemEval with gpt-5-mini, 84.23% with the official gpt-4o (highest openly reproducible score per their post).
Technical surface
- API surface
- SDK: TypeScript
- Backend storage
- pluggable (Postgres, LibSQL, etc.)
- Deployment
- Both
- Embedding model
- multiple supported
- Multi-tenancy
- namespace (resourceId, threadId)
- MCP
- native (first-party) — Mastra MCP server + client
- A2A
- supported — Mastra A2A protocol
- OpenTelemetry
- first-class — built-in OTel support
Alternatives to Mastra Memory → live options ranked by shared design, licence, hosting and MCP support
Similar systems
Other framework-embedded memory in the catalog, ranked by inbound references.
- Spring AI ChatMemory T3
Wraps conversation memory behind an Advisor abstraction. MessageWindowChatMemory default 20-msg sliding window (preserves system messages); pluggable ChatMemoryRepository backends (in-memory, JDBC, Cassandra, Neo4j). Memory injected as cross-cutting concern at the ChatClient level.
- Flowise Memory T1
Discontinued: Flowise (acquired by Workday, Aug 2025) was sunset in 2026 — code freeze 2026-07-29, GitHub repo archived 2026-08-13, end of life 2026-08-31. Was a visual LangChain canvas with LangChain-native memory nodes (Buffer, Buffer Window, Conversation Summary Buffer) backed by an internal chat_message DB table; Buffer Window kept the last K interactions, Summary Buffer compressed older turns while keeping recent ones verbatim.
- LangGraph Persistence T2
Distinct from LangMem. Built-in checkpointer saves graph state per super-step (short-term, thread-scoped). Store System adds long-term hierarchical key-value memory across threads with optional vector search + TTL. Postgres / Mongo / Redis stores for production.
- AutoGen Memory T2
AutoGen's memory layer: ListMemory chronological context plus teachable agents that vectorise corrections; integrates with Mem0/Zep rather than building deep memory natively. Microsoft AutoGen is in maintenance mode (community managed, bug and security fixes only), and Microsoft points new projects to Microsoft Agent Framework.
- CrewAI Memory T2
Memory subsystem inside the CrewAI orchestration framework; integrates with Mem0 for the long-term tier.
- Hermes Agent memory (Nous Research) T3
Nous Research's open-source self-improving agent keeps an agent-curated MEMORY.md/USER.md notebook injected at session start, full-text (FTS5) search over past sessions that returns raw messages without LLM summarisation, and optional Honcho user modelling. one of the most-starred agent projects of 2026.
Sources (65 cited facts, 3 third-party)
Each fact links to the page it came from; the badge says who published it (how to read sources).
- What it is LibSQL + semantic recall vendor
- What it does & what's distinct TypeScript-first agent framework. Default memory: LibSQL store + on-device fastembed-js … vendor
- Claims / benefits "Observational Memory" reports 94.87% on LongMemEval with gpt-5-mini, 84.23% with the of… vendor
- Modalities text vendor
- Created 2025-10 vendor
- Latest release @mastra/core v1.79.0 (2026-10-21) repo
- License Apache-2.0 (core; ee/ dirs under source-available Mastra Enterprise License) repo
- GitHub mastra-ai/mastra — ~27.1k stars, last commit 2026-10 repo
- Mindshare ~1.8M/mo npm (mastra pkg Jan 2 3 press ~15 (small YC-b jobs third-party
- Funding $35M total Series A · 2026-01 vendor
- Customers / scale 11-50 ppl no data (300K+ weekly npm downloads, 22K GitHub stars) vendor
- Pricing Free + paid vendor
- Compliance SOC 2 Type II vendor
- Deployment Both vendor
- HQ US vendor
- Founders / pedigree Sam Bhagwat (CEO, Stanford, ex-Zenefits/PlanGrid); Abhi Aiyer (CTO); Shane Thomas (CPO, … vendor
- Memory volume / scale no scale claim vendor
- Performance 94.87% LMES (Mastra observational memory research, 2025) vendor
- API surface SDK: TypeScript vendor
- Backend storage pluggable (Postgres, LibSQL, etc.) vendor
- Multi-tenancy namespace (resourceId, threadId) vendor
- Embedding model multiple supported vendor
- Versioning immutable append-log vendor
- Tombstoning append-only no delete vendor
- Schema evolution schema with migration vendor
- Namespace primitives per-tenant vendor
- Contradiction handling always append (no resolution) vendor
- Forgetting policy none (permanent) vendor
- MCP support native (first-party) — Mastra MCP server + client vendor
- A2A support supported — Mastra A2A protocol vendor
- OTel first-class — built-in OTel support vendor
- Webhooks / events webhooks supported vendor
- Import / export JSON; LibSQL/Postgres backends vendor
- Integration count 0 repo
- Orchestration both (single Agent + multi-agent workflows; thread/resource scoping) vendor
- Programmatic control full CRUD via TS SDK vendor
- Vendor benchmarks LMES 94.87% with gpt-5-mini; 84.23% with gpt-4o (vendor) vendor
- Pricing specifics OSS free + Mastra Cloud paid (price tiers not detailed publicly) vendor
- Agent abstraction DAG (Workflow) + actor (Agent) vendor
- Memory primitives multiple (LibSQL key-value + vector via fastembed-js) vendor
- LLM lock any (BYO; OpenAI/Anthropic/Google/local) vendor
- Runtimes Node/TypeScript vendor
- Session handling persistent thread (per-thread memory) vendor
- Validated verticals Node.js/TypeScript agent developers vendor
- Time to running <10min (npm install @mastra/core) vendor
- Anti-fit not for Python-only stacks vendor
- Optimised for TypeScript-native developer experience + LibSQL semantic recall vendor
- Adjacent infrastructure requires Mastra; BYO LLM; LibSQL bundled vendor
- Pros Modern TypeScript-first framework with memory built in; LibSQL default substrate keeps o… vendor
- Cons TypeScript ecosystem — less appealing for Python-only shops. vendor
- Project & sources Mastra Memory docs Observational Memory research agent memory guide vendor
- download-trajectory [{"ym":"2021-12","monthly":0,"cum":0},{"ym":"2022-01","monthly":0,"cum":0},{"ym":"2022-0… registry
- cost-input-usd-per-mtok 0 vendor
- cost-output-usd-per-mtok 0 vendor
- cost-tier free vendor
- cost-pricing-model free-tier vendor
- cost-last-verified 2026-06-29 vendor
- adoption-accelerator yc-w25 third-party
- architecture-deployment-locus hybrid vendor
- licensing-bucket open-core repo
- positioning-platform-vs-product platform vendor
- taxonomy-three-layer-rollup agent-infra repo
- developer-activity-opendigger {"activity":1513.71,"as_of":"2026-09","openrank_12mo":109.68,"openrank_current":118.19,"… third-party
- sdk-languages typescript repo
- protocol-support mcp-client,mcp-server vendor
What changed
Changes to this record's lifecycle, licence, release, pricing and hosting facts, from the catalog's git history. Subscribe (Atom)
Related systems
References (1)
- Opencode (SST) integrates with — Mastra repo ships @mastra/opencode, an OpenCode plugin for Mastra Observational Memory. (verified 2026-10-29)