Tenzro Testnet is live —request testnet TNZO
Open call · partner cohort 1

Launch your agent on Tenzro.

Agentic transactions are the 2026 opportunity. Tenzro provides the identity, payment, and settlement primitives an autonomous agent actually needs to transact at scale.

400 ms

Block time

~3,570 tx/s

Theoretical max

5 M tx/day

Phase A target

15 M tx/day

Phase B target

Program structure

Three phases.
Sequenced for scale.

Tenzro Labs runs the reference cohort first to harden the deployment path, then opens to vetted partners, then to anyone. Each phase must hit its targets before the next opens.

Phase AActive

Tenzro reference cohort

100+ agents operated by Tenzro Labs across 5 fleet types — payment routing, bridge arbitrage, multi-modal inference, yield rebalancing, RWA + Canton settlement.

Phase BApplications open

Open-call partner cohort

Approved partners receive a SeedAgent Charter with decaying treasury allocation, dashboard listing, SDK access, and co-marketing.

Phase CComing

Permissionless deploy

Self-serve agent registration via the web launchpad once Phases A and B have validated the deployment path.

What partners receive

The Charter package.

Every Phase B partner receives a governance-approved Charter drawn from the SeedAgent treasury earmark — not the public faucet pool.

Treasury allocation

Per-Charter SpendCaps drawn from a SeedAgent earmark — separate from the public faucet pool.

Decay schedule

100% for months 0–2, then 75 → 50 → 25 → 0 over the following nine months. Anti-camping by design.

SDK access

Rust + TypeScript SDKs, agent-kit templates, and the same MCP and A2A endpoints Tenzro reference agents use.

Public dashboard

Your agent's identity, transaction count, and uptime listed on the Tenzro launchpad.

Co-marketing

Featured in launch announcements, ecosystem updates, and the partner directory.

Direct support

Channel into Tenzro engineering for deployment help, SDK gaps, and template extension.

Reference templates

Eighteen templates.
Fork one. Ship.

Every template in tenzro-agent-kit is a real, end-to-end agent with identity, wallet, capabilities, and the Tenzro stack pre-wired. Fifteen run on CPU-only nodes today.

TemplateCategoryModalityStatus
Intelligent Payment RouterPaymentsLogicAvailable
MPP Payment AgentPaymentsLogicAvailable
Bridge Arbitrage ScannerCross-chainLogicAvailable
Cross-chain Liquidity AggregatorCross-chainLogicAvailable
Yield RebalancerDeFiLogicAvailable
Multi-chain Portfolio ManagerDeFiLogicAvailable
Autonomous RWA CustodianTokenizationLogicAvailable
Canton Trade SettlerEnterpriseLogicAvailable
Model Inference ProxyAIMulti-modalAvailable
Agentic Inference MarketplaceAIMulti-modalAvailable
Audio TranscriberAIAudio (ASR)Available
Vision IndexerAIVisionAvailable
Timeseries ForecasterAITimeseriesAvailable
Premium Alpha AdvisorAIVision + TimeseriesAvailable
Video AnalystAIVideoAvailable
Language TrainerTrainingText — GPUPending training infra
Vision TrainerTrainingVision — GPUPending training infra
Timeseries TrainerTrainingTimeseries — GPUPending training infra
Compute

CPU-runnable today.

Tenzro ships seven ONNX runtimes wired into the inference router. The models below run on commodity CPU nodes with permissive or already-gated licenses.

Text embedding
EmbeddingGemma-300M, BGE-M3, Snowflake Arctic L v2.0, Qwen3-Embedding 0.6B
Vision
CLIP ViT-B/32, SigLIP2 base, DINOv3 vits16/vitb16
Timeseries
Granite-TTM-r2, Chronos-Bolt small/base, TimesFM 2.5
Detection
RF-DETR nano/small, D-FINE n/s
Segmentation
EdgeSAM, MobileSAM
Audio (ASR)
Moonshine v2 tiny/base, Distil-Whisper small.en
Testnet capacity

Volume projections.

Verified against live consensus and VM configuration. Projections stay below 5% of theoretical capacity — comfortable headroom for Phase A and Phase B combined.

400 ms

Block time

consensus config

30 M

Max gas/block

VM config

~308 M tx/day

Theoretical ceiling

30M / 21k × 86,400 / 0.4

15 M tx/day

Combined target

Phase A + B steady-state

Notional GMV-equivalent: $50–200 M / month projected at combined steady-state, bounded by per-Charter SpendCaps. Testnet TNZO has no monetary value — these figures are projections to validate the protocol at scale before mainnet.
How to apply

Apply for a Charter.

Email partners@tenzro.network with the items below. The Tenzro Labs team reviews applications on a rolling basis.

  1. 01

    Team

    Who is building. Background, links, prior agent or distributed-systems work.

  2. 02

    Agent concept

    What your agent does. Map it to one or more reference templates, or describe the novel template you'd add.

  3. 03

    Target rails

    Which payment protocols (MPP, x402, AP2, Visa TAP, Mastercard, Tempo) and bridges (Wormhole, LayerZero, CCIP, deBridge, Canton, Li.Fi) you'll exercise.

  4. 04

    Projected volume

    Tx/day target and the operations breakdown. Volume is negotiated per Charter.

  5. 05

    Telemetry plan

    How your agent will report uptime, transaction counts, and failure modes into the partner dashboard.

FAQ

Common questions.

Can agents use the public faucet?

Yes. The faucet is address-based, not identity-gated. An agent calls tenzro_faucet with its auto-provisioned MPC wallet address. Default 100 TNZO per request, max 1,000 TNZO. The faucet alone is too small for a sustained agent fleet — that's what Charters are for.

What's a Charter?

A SeedAgent Charter is a governance-approved mandate that draws from a treasury earmark separate from the public faucet pool. It carries SpendCaps, an allowed-operations whitelist, a counterparty filter, and a sunset date. Allocations decay on a fixed schedule so partners can't camp on funding.

Which templates can I run today without GPUs?

15 of the 18 reference templates run on commodity CPU nodes. The three trainers (language, vision, timeseries) need GPU and are gated until Phase 9 training infrastructure is live.

How is partner volume distinguished from Tenzro-operated volume?

The public dashboard separates the two cohorts. Partner agents register their own DIDs and the block explorer attributes transactions accordingly. We don't conflate them.

What's the application bar for Phase B?

Demonstrated technical capability to operate an autonomous agent fleet, a clear use case mapped to one or more reference templates or a credible novel template, and the ability to provide telemetry into the partner dashboard. Volume targets are negotiated per Charter.

Is this mainnet or testnet?

Testnet. Tokens are testnet TNZO with no monetary value. Volume and GMV-equivalent figures are projections to validate the protocol at scale before mainnet.