Statewave is live · free · last verified 5 Nov 2026 · viability healthy
Statewave
Open-source AI agent memory runtime that runs on your own Postgres and issues a signed receipt for every answer.
The signed, replayable receipt is the actual product here — when an audit or a bad answer lands, most memory layers can't reconstruct what the agent was told, and this one can by default. Labels, YAML policy gates, log_only testing and one-call subject erasure back the claim with real surface area, not slideware. What you pay is operational: a Postgres, a runtime and background compilation are yours to run.
Verdict generatedVerified 8d ago · liveness 76/100 · cite: rightaichoice.com/tools/statewave
- Engineering teams building AI agents that must remember users, accounts and prior sessions across restarts
- Support engineering teams needing session-aware context, handoff packs and repeat-issue detection
- Platform teams already running self-hosted Postgres who want the memory boundary inside their network
- Compliance-driven orgs needing provenance, replayable receipts and one-call GDPR subject erasure
- Non-technical users who want a hosted, plug-and-play memory service
- Teams unwilling to self-host and operate a Postgres-backed runtime
- Projects that only need simple key-value memory without compilation or governance
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Statewave if you want agent memory as a managed endpoint and are not prepared to run Postgres, background compilation and policy operations yourself.
Self-hosting means you carry the compute and storage bill for Postgres plus the compilation pass, and that grows with the number of episodes you retain rather than with seat count.
Statewave is free under Apache 2.0 with no tier list, so the real cost comparison is operational: versus managed memory APIs like Mem0 or Zep you trade a subscription for Postgres hosting, compilation compute and the engineer hours to run them. For a platform team that already operates Postgres, that trade is usually favorable at scale; for a two-person startup wanting an endpoint today, a hosted service is cheaper in total cost even with a monthly bill.
In short
Statewave — Open-source AI agent memory runtime that runs on your own Postgres and issues a signed receipt for every answer. Best for Engineering teams building AI agents that must remember users, accounts and prior sessions across restarts, Support engineering teams needing session-aware context, handoff packs and repeat-issue detection, Platform teams already running self-hosted Postgres who want the memory boundary inside their network. Free to use.
What is Statewave?
Statewave — Open-source AI agent memory runtime that runs on your own Postgres and issues a signed receipt for every answer. Best for Engineering teams building AI agents that must remember users, accounts and prior sessions across restarts, Support engineering teams needing session-aware context, handoff packs and repeat-issue detection, Platform teams already running self-hosted Postgres who want the memory boundary inside their network. Free to use.
Statewave is a self-hosted memory runtime for AI agents, published under Apache-2.0 and designed to sit between your agent and your own Postgres + pgvector instance. Instead of bolting a vector store onto each app, you get a small three-call surface: record every event as an immutable episode, compile those episodes into typed memories with confidence scores and validity windows, then retrieve a ranked context bundle trimmed to your token budget. Pass emit_receipt: true on the retrieval call and the same response includes a signed, replayable record of exactly what the agent saw. The governance layer is the part most memory APIs skip. Memories carry sensitivity labels (pii, financial, secret) that heuristic detectors suggest at compile time and an operator promotes. YAML policies deny or redact per caller and tenant, with a log_only mode so you can watch enforcement before turning on denies. Subjects — users, accounts, agents, repos — organize everything, tenant isolation is scoped in the admin console, and subject-level erasure wipes episodes, memories and receipts in one call. The vendor benchmarks on a fork of Mem0's harness with judge and scoring code untouched: LoCoMo n=1,540 puts Statewave at 0.905 versus Mem0 cloud 0.899 and Mem0 OSS 0.866. That's a real lead over the open-source tier and inside single-run noise against the paid one — the vendor says so itself. Setup is one npx command that boots the API, admin console and Postgres locally with no account and no telemetry path back to the vendor. The MCP server auto-wires Claude Code, Claude Desktop, Cursor, VS Code Copilot and Codex CLI, while connectors feed GitHub, Slack, Zendesk, Jira, Notion, Intercom, Gmail and Discord in as episodes. It competes with managed memory endpoints like Mem0 and Zep by trading operational weight for audit-grade control over the memory data boundary.
Key features
- Record raw events as immutable, append-only episodes via POST /v1/episodes
- Compile episodes into typed memories with confidence scores and validity windows
- Retrieve ranked, token-bounded context bundles with POST /v1/context
- Emit signed, replayable state-assembly receipts with emit_receipt: true
- Trace every compiled memory back to its source episodes
- Auto-suggest pii, financial and secret sensitivity labels at compile time
- Gate memory access with YAML policies per caller and tenant
- Audit policy enforcement in log_only mode before enabling denies
- GDPR-ready subject erasure across episodes, memories and receipts in one call
- Multi-tenant isolation scoped per subject and tenant in the admin console
- Auto-wire MCP servers for Claude Code, Claude Desktop, Cursor, VS Code Copilot, Codex CLI
- REST API with Python and TypeScript SDKs
- OpenRouter proxy for persistent memory on OpenRouter calls
- Run locally with one npx command, no account, no telemetry path to vendor
- Fully local default compiler; optional external LLM provider with your own key
Behind the verdict
Pick Statewave when the question "what did the agent know when it answered?" has to have an answer you can hand to a reviewer. Support desks, coding assistants and multi-agent systems all read from the same subject-keyed memory, and the receipt turns each retrieval into evidence rather than a log line you hope is complete. The governance story is unusually concrete for an open-source memory tool. Sensitivity labels get suggested at compile time, then an operator promotes them; policies are YAML with a log_only mode so you can watch what would have been redacted before you enforce anything. Subject-level erasure covering episodes, memories and receipts in one call is the kind of thing GDPR requests actually demand. Where it bites is operations. You're running Postgres with pgvector, a Python 3.11+ runtime and background compilation. The quickstart is a single npx command with no account, but production is your Helm chart or Compose file, your backups, your upgrade cadence. Teams that want memory as a managed endpoint should look at Mem0 or Zep instead — shorter path, less control. On quality, the LoCoMo numbers (0.905 vs 0.899 Mem0 cloud, 0.866 Mem0 OSS on n=1,540) are worth reading carefully. Beating the open-source baseline by roughly four points is meaningful; matching the paid tier is parity within noise. Neither number tells you how retrieval ranks against your own traffic, so budget time to tune against real episodes before you trust it in front of customers. The compiler choice matters for data flow. The default compiler runs fully local, so with no LLM key configured nothing leaves your network; point it at a provider and that calculus changes. Worth deciding deliberately rather than by default. MCP wiring is a genuine convenience — Claude Code, Claude
Limitations
- Statewave is self-hosted Apache-2.0 software you run on your own Postgres, so operating the runtime, background compilation pass and governance controls is your responsibility.
- The evidence shows only API calls (POST /v1/episodes, /v1/memories/compile, /v1/context) plus Python and TypeScript SDKs and an npx runner as integration channels.
- Governance features (sensitivity labels, YAML policies, signed receipts, tenant isolation, subject erasure) are vendor-described; the evidence provides no independent validation of these security guarantees.
as of 2026-10-18
Researching Statewave? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Is Statewave still active?
Statewave is live · free · last verified 5 Nov 2026 · viability healthy.
Viability score
How well maintained and how widely used is Statewave? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: November 2026
How we score →What's new in Statewave
Checked 6 days agoAcross the latest 4 updates: 4 community discussions.
Statewave Guide — Day 5: We taught it to speak. Then realized it still doesn't finish the job
Statewave said guidance compilation fixed the language, but the system still stopped halfway through the task. Correct English is not the same as complete guidance.
Statewave Guide — Day 4: We passed the safety test. Then failed the product test
Statewave reported nothing dangerously wrong but little useful output — and said the cause was not the missing capabilities it expected.
Statewave Guide — Day 3: True is not the same as relevant
Statewave put a real model behind the Guide; it still produced a claim true about the application but wrong about the feature asked about.
Statewave Guide — Day 2: Unknown is better than wrong
Statewave traced a real user action from a button to the backend, discovering its own benchmark was wrong and fixing the benchmark, not the indexer.
What do users say about Statewave?
What people actually say about Statewave — is it worth it?
Our community sentiment research written Aug 13, 2026 is more than 90 days old, so it is being re-reviewed against current data and is not shown.
How we verified this
Every listed fact about Statewave — pricing, features, links, status — is re-read from the vendor's own pages by our verification pipeline and each change is logged with a date; the last pass ran on 5 Nov 2026. Analysis is drafted with AI models and checked against our editorial standards before publishing; see how we verify.
Verification history
We have re-verified Statewave 10 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — changed pricing page, limitations
- — changed SSL certificate
- — changed summary, pricing tiers, integrations, our verdict, features, description
- — changed pricing page, platforms, limitations
- — changed who it suits, integrations, description, pricing tiers, features, summary, who should skip it, our verdict
- — changed pricing page
Showing the 6 most recent of 10 verification passes.
Free to cite with attribution — this page re-verifies continuously.
How much does Statewave cost in 2026?
Statewave: Free to use. Pricing last verified 5 Nov 2026.
Plans compared
Every published Statewave tier, its price as the vendor publishes it, who it fits, and what it adds over the tier below. The 12-month cost projection follows.
| Plan | Price | Billing | Who it's for |
|---|---|---|---|
| Open Source (Apache 2.0) | Free | — | Platform and support engineering teams who already run Postgres and need governed, auditable agent memory they own end to endAdds: Free entry point with the full runtime — ingest, compile, retrieve, admin console, YAML policies, sensitivity labels, state-assembly receipts, SDKs and MCP auto-configuration — self-hosted under Apache 2.0 |
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Where the pricing makes sense
The company stage and team size where Statewave's pricing actually pencils out — and where peers do it cheaper.
Statewave is free under Apache 2.0 with no tier list, so the real cost comparison is operational: versus managed memory APIs like Mem0 or Zep you trade a subscription for Postgres hosting, compilation compute and the engineer hours to run them. For a platform team that already operates Postgres, that trade is usually favorable at scale; for a two-person startup wanting an endpoint today, a hosted service is cheaper in total cost even with a monthly bill.
Has Statewave's pricing changed?
No plan price for Statewave has moved since we began logging in Aug 2026; the pricing page is re-read on every verification pass.
Who should pick Statewave?
Pick Statewave if you are Engineering teams building AI agents that must remember users, accounts and prior sessions across restarts, Support engineering teams needing session-aware context, handoff packs and repeat-issue detection, Platform teams already running self-hosted Postgres who want the memory boundary inside their network, Compliance-driven orgs needing provenance, replayable receipts and one-call GDPR subject erasure.
Look elsewhere if you are Non-technical users who want a hosted, plug-and-play memory service, Teams unwilling to self-host and operate a Postgres-backed runtime, Projects that only need simple key-value memory without compilation or governance, Buyers who need memory live this week with no tuning of retrieval relevance against their own traffic.
Skip it if Skip Statewave if you want agent memory as a managed endpoint and are not prepared to run Postgres, background compilation and policy operations yourself.
GeneratedReal-world workflow fit
Concrete scenarios for the personas Statewave actually fits — and what changes day-one when you adopt it.
You point the Zendesk and Slack connectors at Statewave so closed tickets and thread replies land as episodes, let compilation produce typed prior-issue summaries, and assemble handoff packs that fit a fixed token budget for the live queue.
Outcome: An agent picking up a conversation gets ranked context about the account's history without replaying full transcripts, and every pack carries a receipt you can replay when a customer disputes what was promised.
You boot Statewave with the one-line npx command against your own Postgres, enable log_only policy mode, and run a month of traffic to see which callers would have been denied before turning enforcement on.
Outcome: You get an evidence base for the YAML policy rules instead of guessing, and a log of enforcement decisions to show reviewers before anything is blocked in production.
You wire Statewave into Claude Code and Cursor via MCP auto-configuration so architecture decisions, tech-stack choices and preferences accumulate as memories against the repo subject across sessions.
Outcome: The agent stops re-deriving project context each session, and you can trace any remembered decision back to the episode where it was made.
Use cases
- Build AI support agents that carry session context, resolution history and handoff packs across conversations
- Run coding agents that accumulate project knowledge — tech stack, architecture decisions, preferences — over time
- Give internal copilots persistent memory of user workflows, past decisions and organizational context
- Assemble handoff context packs that fit a fixed token budget for a live support queue
- Keep an audit trail of what each agent saw via immutable state-assembly receipts
- Satisfy GDPR erasure requests by wiping episodes, memories and receipts for a subject in one API call
- Operate multi-tenant memory with per-tenant policy enforcement and app-layer query isolation
- Feed GitHub and Zendesk events in as normalized episodes for a support or engineering agent
Setup time & first value
How long it actually takes to get something useful out of Statewave — broken out by persona, not the marketing-page minute.
Local boot to a running API, admin console and Postgres is one npx command with no account, and the vendor's manual docker compose path is documented at about five minutes. MCP clients auto-wire from that same command. Time to first useful memory depends on your connector setup — budget an afternoon for GitHub or Zendesk ingestion plus your first YAML policy in log_only mode.
Switching to or from Statewave
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From Mem0 or Zep: export stored memories and replay them as episodes through the ingest API, then run compilation so Statewave derives its own typed memories with provenance.
- →From prompt-stuffed chat history: post existing transcripts as episodes per subject, let the compile pass extract profile facts and prior-issue summaries, and drop the full-history block from your prompt.
- →From a raw vector store: keep it running in parallel while you feed the same events to Statewave, and compare ranked token-bounded bundles against nearest-neighbor chunks before cutting over.
- ↗To Mem0 or Zep: retrieve compiled memories via the REST API and push them into the managed service's memory endpoints; state-assembly receipts do not transfer and would need to be archived separately.
- ↗To a plain vector store: export memories with their source episode links, but expect to lose confidence scores, sensitivity labels and the policy enforcement layer.
Integrations
Resources & guides
- Documentationstatewave.ai
Docs · Statewave
Full product docs from statewave.ai
- Resourcegithub.com
Statewave Docs · Statewave
Helpful link from github.com
- Quickstartgithub.com
Getting Started · Statewave
Get up and running fast from github.com
- Resourcegithub.com
Troubleshooting · Statewave
Helpful link from github.com
- Resourcegithub.com
Receipts · Statewave
Helpful link from github.com
- Resourcegithub.com
Sensitivity Labels · Statewave
Helpful link from github.com
- Resourcegithub.com
Subject Design · Statewave
Helpful link from github.com
- Resourcegithub.com
Why Statewave · Statewave
Helpful link from github.com
- API Referencegithub.com
Api · Statewave
Methods, params, types from github.com
- Resourcegithub.com
Connectors · Statewave
Helpful link from github.com
Tutorials & Learning
YouTube returned 6 videos for “Statewave”, and we withheld 6: 6 could not be judged, because “Statewave” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Statewave.
Official links
What are the best Statewave alternatives?
The closest alternatives to Statewave we list are Mem0, Versuno AI and Cognee. Ranked by direct product-type match, not category overlap.
| Alternative | Pricing | Starts at | Last verified |
|---|---|---|---|
| Mem0 | Freemium | Pricing not published | — |
| Versuno AI | Freemium | Pricing not published | — |
| Cognee | Freemium | Pricing not published | — |
Alternatives to Statewave
View all
Mem0
Mem0 is an AI memory layer that gives agents persistent context across sessions, users and apps
Versuno AI
Open-source Postgres memory engine for AI agents that sends memory conflicts to a human instead of overwriting them.
Tools that pair well with Statewave
Common stack mates teams adopt alongside Statewave, with the specific reason each pairing earns its keep.
Mem0
Mem0 is an AI memory layer that gives agents persistent context across sessions, users and apps
Versuno AI
Open-source Postgres memory engine for AI agents that sends memory conflicts to a human instead of overwriting them.
Cognee
Cognee is open-source agent memory that turns scattered company data into a cited knowledge graph agents can recall.
Statewave vs its closest rival
Statewave's closest head-to-head is with Mem0 — If you need a battle-tested memory layer with compliance (HIPAA, SOC 2) and deep framework integrations, Mem0 is the clear choice—especially for regulated industries. But if you're a startup that values open-source flexibility and want to self-host without vendor lock-in, Statewave offers a lean, customizable runtime. Choose based on whether you prioritize compliance and ecosystem (Mem0) or open-source control (Statewave). Verdict generated
Head-to-head comparisons
Statewave vs Mem0
If you need a battle-tested memory layer with compliance (HIPAA, SOC 2) and deep framework integrations, Mem0 is the clear choice—especially for regulated industries. But if you're a startup that values open-source flexibility and want to self-host without vendor lock-in, Statewave offers a lean, customizable runtime. Choose based on whether you prioritize compliance and ecosystem (Mem0) or open-source control (Statewave).
GeneratedStatewave vs Tobira
If you're building agents that need to be found and communicated with by humans and other agents across the open web, start with Tobira — it's free and gives you instant public identity. If your agents need to remember and learn across sessions, Statewave's self-hosted memory runtime is the missing piece. Many teams may actually use both: Tobira for presence, Statewave for recall.
GeneratedStatewave vs Arcade Ai
Pick Statewave if your bottleneck is agents forgetting context and you want to self-host a memory layer you fully control. Pick Arcade AI if your bottleneck is securely connecting agents to real user accounts and enterprise tools — it ships auth, governance, and a huge MCP catalog out of the box. Most teams shipping production agents today will find Arcade's faster time-to-value worth the trade-off, unless you have a very specific memory-heavy use case.
GeneratedFrequently asked questions
Categories
Topics
For makers of Statewave
- Get a Verified Listing for Statewave ($39): a written decision within 1 working day.
- Free listing for another tool: Free reviews are decided in order, within 3 weeks of the review starting.
Used Statewave? Help shape our editorial sentiment research.