The agent interaction layer for tokenized assets
Silvana enables AI agents, applications, and institutions to securely operate tokenized assets on Canton — deploy agents that trade, settle, and prove.
Trading takes minutes. Your first agent takes an afternoon.
- <1s
- Order matching
- 100%
- Atomic settlement
- 1,240+
- Agents running
One platform, every agentic workflow.
A complete set of trading and agent infrastructure — explore each on its own, or compose them together.
Trade privately on a high-performance orderbook.
Off-chain matching in under a second. Zero information leakage, zero slippage, zero pools. Your orders never touch a public mempool.
Explore Silvana BookPut agents on the flow.
Market making, grid, RFQ, settlement, proving — configure an agent once and let it run around the clock.
Browse Agent SpaceSwap with atomic settlement.
Assets and payment move simultaneously — or not at all. No pools, no custody handoffs, rollback-protected.
Try Swap
CC▾0Build on the Agentic API.
One gRPC interface for everything on Canton: payments, DvP, transfers, multicall workflows.
Read the API referenceProve without revealing.
Hand auditors ZK proofs of trading activity — without exposing a single order.
How proving worksEarn as you trade.
Every settled transaction counts toward seasonal rewards in EarnHub.
Open EarnHubNot sure where to start?
Tell us what you operate — a desk, a treasury, a product, a strategy — and we’ll point you to the right rails.
Built for production volume.
order matching, off-chain and optimistic
atomic settlement: every trade settles in full or rolls back
live markets, with more launching every season
agents deployed and running
Trade size without showing your hand.
Run OTC and block flows through private RFQ. Settle atomically via DvP on Canton — no counterparty risk, no partial states. And when reporting season comes, hand over ZK proofs instead of spreadsheets: selective disclosure means full accountability with zero exposure.
Confidentiality, deterministic execution, and compliance-ready proofs — infrastructure that fits how institutions actually operate.
From strategy to running agent in an afternoon.
A general-purpose SDK, proto interfaces for any language, and four hosting models — from your laptop to a TEE. Onboard with one CLI command, configure your strategy in agent.toml, and run grid, market-making, or taker flows out of the box.
$ silvana onboard --name my-agent --invite EARLYBIRD
✓ Key generated · .env written · agent.toml created
$ silvana agent run
→ Connected to CC/USDC · quoting 6 levels · streaming settlement events…Quote both sides. Keep your book to yourself.
Grid orders and RFQ provision on private rails — your inventory, spreads, and pricing logic stay invisible to the market. Configure depth and risk once; agents keep quotes fresh while you sleep. Every settled fill counts toward EarnHub rewards.
“Silvana gives our users something wallets alone never could: assets that stay under their control while agents do the work.”
Reliable, extensible infrastructure for every stack.
Adapt Silvana to your architecture, not the other way around.
Don't code?
Trade from the Terminal, swap in two clicks, and track rewards in EarnHub — no integration required.
Open the appConfigure an agent.
Simulate a strategy in the Playground, then export the exact agent.toml and CLI command to run it for real.
Try the PlaygroundBuild your own.
gRPC services defined in proto work from any language. Agents run in Rust; a TypeScript SDK covers web. Two-phase signing keeps keys local.
Read the docsScale with confidence: Sub-second matching, bidirectional settlement streams, and canonical update IDs on every transaction — so your systems reconcile on the first pass, even at volume.
The latest from Silvana.
The agentic economy, in your inbox.
Product launches, new markets, and deep dives on agent infrastructure. Once or twice a month. No noise.
Unsubscribe anytime. We never share your email.
Ready to put agents to work?
Create an account and trade in minutes, or talk to us about what you're building.