Statewave vs Tobira

Side-by-side comparison of features, pricing, and ratings

Analysis reviewed Live tool data as of 2026-09-28
Cross-checked through our multi-step verification ·
Saved

At a glance

DimensionStatewaveTobira
PricingFreemiumFree
Primary FocusMemory runtime for agentsAgent identity & discovery
Key FeaturesPersistent memory, semantic search, observability@handles, profiles, guest chat, multiple discovery files
DeploymentSelf-hosted (open-source)Open network (decentralized)
Best ForDevelopers needing scalable memory for production agentsWeb3 developers, early adopters claiming agent identity
Recent NewsNo recent news capturedPublic launch of core primitives (2026-06-28)

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.

Statewave
Statewave

Open-source, self-hosted memory runtime that gives AI agents governed long-term memory with provenance and audit receipts

Visit Website
Tobira
Tobira

Tobira gives AI agents public addresses—@handles, profiles, and guest chat for agent identity.

Visit Website
Pricing
Free
Free
Plans
$0
$0/mo
Popularity
3 views
5.7k views
Skill Level
Advanced
Beginner-friendly
API Available
Platforms
WebAPICLI
Web
Categories
🧠 Agent Memory & Runtimes
🧠 Agent Memory & Runtimes🔌 MCP Servers & Agent Tooling
Features
Ingest raw events as immutable episodic memories
Compile episodes into typed memories with confidence scores, idempotent on re-run
Retrieve ranked, token-bounded context bundles at 512, 1024, 2048 or 4096 tokens
Trace every compiled memory back to its source episodes with full provenance
Tag memories with pii, financial or secret sensitivity labels, auto-detected at ingest
Gate memory access with declarative YAML policies per tenant and caller identity
Emit immutable ULID-addressable state-assembly receipts with a byte-level integrity hash
Multi-tenant isolation via subject-scoped architecture and app-layer query isolation
GDPR-ready subject erasure across episodes, memories and receipts in one call
Organize memory around subjects: users, accounts, agents, repos
Auto-configure MCP servers for Claude Code, Claude Desktop, Cursor, VS Code Copilot and Codex CLI
Support workflow: session-aware context, handoff packs, customer health scoring, SLA tracking, repeat-issue detection
Run locally with one npx command, no account, offline
REST API with Python and TypeScript SDKs
log_only policy mode to audit enforcement decisions before enabling them
Claim a public @handle as an AI agent address
Publish a profile with offers, needs, goals, blockers, and proof
Guest chat for human-to-agent conversation
agent.json machine-readable discovery file
guest-agent.json for visiting-AI access
llms.txt for AI assistants
AGENTS.md discovery file for agents
Link headers and discovery metadata for agent lookup
Site Agent representing a website or business
Attached Knowledge for approved agent context
Channel instructions to route agent conversations
Permissionless, censorship-resistant registration
Profile pages as public address surfaces
Integrations
GitHub
Slack
Zapier
n8n
Zendesk
Claude Code
Claude Desktop
Cursor
VS Code Copilot
Codex CLI
Cline
Continue
Windsurf
Zed
Aider

Who should pick which

  • Web3 developer building decentralized agent infra
    Pick: Tobira

    Tobira is built for this — decentralised identity, agent.json, and permissionless interaction align with Web3 ethos.

  • Solo AI agent developer needing persistent context
    Pick: Statewave

    Statewave's self-hosted, open-source memory runtime lets you add long-term recall without depending on a third-party cloud.

  • Enterprise team deploying agents at scale
    Pick: Statewave

    Statewave offers observability, high availability, and scalability — enterprise needs that Tobira doesn't address.

  • Researcher exploring agent identity and discovery
    Pick: Tobira

    Tobira's multiple discovery surfaces (llms.txt, AGENTS.md) are perfect for studying how agents are found.

  • Startup wanting both identity and memory
    Pick: Statewave

    While Tobira is free, Statewave is more critical for product functionality — you can add manual discovery elsewhere, but memory is harder to build from scratch.

Frequently Asked Questions

Statewave vs Tobira: which should you choose?

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.

Can I use Tobira without knowing Web3?

Tobira is designed for Web3 developers, so expect decentralized concepts like permissionless and agent.json. Non-Web3 devs may find the learning curve steep.

Is Statewave fully managed?

No, it's self-hosted open-source. Teams wanting a managed cloud service will need to wait — the data notes that as not available now.

Does Tobira provide memory for my agent?

No. Tobira focuses on identity and discovery, not memory. You'd need a separate memory runtime like Statewave for that.

What kind of support do these tools offer?

The data doesn't mention support options — both are early-stage infrastructure projects likely relying on community support.

Can I claim a @handle for my business website on Tobira?

Yes, Tobira supports Site Agent for representing websites, so you can claim a handle for a business presence.

Does Statewave integrate with my existing agent framework?

It provides an API for memory read/write, so you can likely integrate it, but no specific framework integrations are listed.

More Statewave or Tobira comparisons

Explore each tool further

Browse these categories

Still deciding? Get the weekly AI tools brief

One email a week — new tools, honest comparisons, no spam.

Last reviewed: August 13, 2026