Tempo, the payments-focused Layer 1 blockchain incubated by Stripe and Paradigm, launched its mainnet on March 18, 2026, alongside the Machine Payments Protocol (MPP), an open standard for machine-to-machine payments co-authored with Stripe. The chain went live with more than 100 integrated servi...
"We look at MPP as another way that you can have a very clear, defined protocol around how an agent communicates with merchants." — Cuy Sheffield, Head of Crypto, Visa
Tempo, the payments-focused Layer 1 blockchain incubated by Stripe and Paradigm, launched its mainnet on March 18, 2026, alongside the Machine Payments Protocol (MPP), an open standard for machine-to-machine payments co-authored with Stripe. The chain went live with more than 100 integrated service providers, design partners including Visa, Mastercard, Deutsche Bank, Standard Chartered, OpenAI, Anthropic, and Shopify, and KlarnaUSD — the first bank-issued stablecoin built on the network.
Tempo raised $500 million in its October 2025 Series A led by Thrive Capital and Greenoaks at a $5 billion valuation. Its mainnet benchmarks 20,000 TPS with a stated path to 200,000+, sub-second finality, and fees below $0.001. The chain has no native gas token; users pay fees in any major stablecoin via an enshrined AMM. The launch represents the first time a company of Stripe's scale — processing over $1.4 trillion in annual payment volume as of 2024 — has deployed its own blockchain to production.
The timing coincides with a broader infrastructure race. Coinbase and Cloudflare's x402 protocol has surpassed 15 million transactions since its September 2025 launch. Visa Crypto Labs shipped its own CLI tool for agent payments on March 18. McKinsey projects AI-agent-driven commerce will reach $3–5 trillion by 2030. Three competing protocols are now live, each backed by a payments incumbent, each targeting the same fee pool.
Tempo's technical choices reflect a single design priority: payment settlement at scale. Built on the Reth SDK using Simplex Consensus (developed by Commonware, which received a $25 million strategic investment from Tempo in November 2025), the chain targets sub-second deterministic finality under normal network conditions.
Testnet benchmarks showed 20,000 TPS. The team claims a path to 200,000+ TPS, which would put Tempo's theoretical throughput in the same range as Visa's stated capacity of 65,000 TPS and Mastercard's 5,000 TPS. Transaction fees target sub-$0.001 per transfer.
The most structurally significant design decision: Tempo has no native gas token. Unlike Ethereum (ETH), Solana (SOL), or nearly every other Layer 1, Tempo validators are compensated in stablecoins. Users pay fees in USDC, USDT, or any other supported stablecoin through an enshrined AMM that converts tokens automatically. This eliminates gas fee volatility — a persistent barrier to blockchain adoption for payment use cases where cost predictability is non-negotiable.
The TIP-20 token standard extends ERC-20 with payment-specific features: batched transfers for payroll and settlement, scheduled payments with protocol-level time windows, and fee sponsorship that allows applications to absorb gas costs on behalf of users. TIP-403 provides an on-chain compliance policy registry shared across multiple tokens.
The chain is EVM-compatible and uses WebAuthn/P256 passkeys for modern authentication, enabling biometric sign-in without seed phrases.
Engineering leadership includes Liam Horne (former CEO, Optimism Labs), Georgios Konstantopoulos (CTO, Paradigm), and Dankrad Feist (former Ethereum Foundation researcher, co-architect of danksharding). Farcaster co-founders Dan Romero and Varun Srinivasan joined the team following Tempo's acquisition of the Ithaca startup. Headcount grew from approximately five employees in August 2025 to 40–50 by launch.
MPP defines a standardized method for software agents and services to request, authorize, and settle payments programmatically. The protocol's designers describe it as "OAuth for money" — authorize once, then allow payments to execute within defined spending limits.
The flow works as follows: an agent requests a resource from a service; the service responds with a payment request specifying amount and denomination; the agent authorizes payment from its wallet; the transaction settles on-chain; and the service delivers the requested resource. MPP introduces a "sessions" primitive that lets agents set a spending cap upfront and stream micropayments continuously, without requiring a separate on-chain transaction for each interaction.
MPP is open-source and designed to be rail-agnostic. While it runs natively on Tempo, the protocol has already been extended to multiple payment networks:
The payments directory launched alongside mainnet catalogs more than 100 MPP-compatible services. At launch, AI agents on Tempo can autonomously purchase compute from Alchemy, query blockchain data from Dune Analytics, and access APIs from Anthropic and OpenAI — all settled in stablecoins without human intervention.
The roster of design partners is the most institutionally dense of any blockchain launch in recent memory:
| Category | Partners | |----------|----------| | Payment Networks | Visa, Mastercard | | Banks | Deutsche Bank, Standard Chartered, Nubank, Lead Bank | | Fintechs | Klarna, Revolut, Ramp, Mercury | | Commerce | Shopify, DoorDash, Coupang | | AI/Tech | OpenAI, Anthropic | | Infrastructure | Alchemy, RedStone, Dune Analytics |
Klarna became the first bank to issue a stablecoin on the network. KlarnaUSD, issued through Bridge's Open Issuance platform, operated on testnet since late 2025 and is expected to launch on mainnet in the near term. Klarna CEO Sebastian Siemiatkowski stated that "crypto infrastructure has reached a point of speed, security and cost efficiency suitable for high-volume payment use cases." The company serves 114 million customers globally.
RedStone was selected as Tempo's core oracle provider after what Tempo described as an "exhaustive evaluation" of the oracle market. The integration provides continuous FX feeds at sub-second update frequency, covering major pairs including USD/KRW and USD/MXN — essential for cross-border settlement pricing.
UBS and Kalshi participated in testnet design from its December 2025 launch.
Tempo is not an isolated product. It is one component of a vertically integrated stablecoin infrastructure stack that Stripe has assembled over the past 18 months:
Bridge ($1.1 billion acquisition, closed February 2025): Stablecoin issuance, orchestration, and reserve management. Bridge received conditional OCC approval for a national trust bank charter, authorizing digital asset custody and stablecoin issuance at the federal level. Post-acquisition, Bridge's transaction volume grew 4x. Customers include Coinbase and SpaceX.
Privy (acquired June 2025): Crypto wallet infrastructure for user onboarding. Eliminates seed phrase friction.
Tempo ($500 million Series A, $5 billion valuation): Settlement layer purpose-built for stablecoin payments.
Open Issuance (via Bridge): A platform for third-party stablecoin creation — used by Klarna for KlarnaUSD.
The combined stack gives Stripe a full circuit: issue stablecoins (Bridge/Open Issuance), onboard users (Privy), settle transactions (Tempo), and monetize flow through its existing $1.4 trillion processing infrastructure. Stripe processes millions of cross-border transactions daily, a segment growing 50% annually, according to the company.
In 2024, stablecoins moved $15.6 trillion in value globally — a figure on par with Visa's annual payment volume. Stripe appears to be positioning itself to capture settlement fees from this flow rather than merely processing card transactions that happen to involve crypto.
Three competing protocols for machine payments are now live, each backed by a payments incumbent:
Stripe/Tempo — Machine Payments Protocol (MPP)
Coinbase/Cloudflare — x402 Protocol
Visa — Visa CLI / Agentic Ready
Visa occupies an unusual position: it contributed to Tempo's MPP specifications while simultaneously shipping its own competing CLI tool. This dual strategy suggests Visa views the machine-payment market as large enough to justify participation on multiple rails simultaneously.
Notably, all three protocols address the same bottleneck: the lack of a standardized way for software to pay for services autonomously. The solutions differ primarily in settlement layer — stablecoin-on-blockchain (Tempo, x402) vs. card-network (Visa CLI) — and in architectural philosophy. x402 is stateless and HTTP-native. MPP introduces sessions and spending caps. Visa CLI extends existing card infrastructure.
Cross-border payments generate an estimated $120 billion in transaction fees annually, according to McKinsey data cited by Klarna. Stablecoin-based settlement offers a structural cost advantage by eliminating correspondent banking intermediaries.
The market opportunity for machine-to-machine payments specifically is nascent but growing. McKinsey projects $3–5 trillion in B2C goods transacted through AI agents by 2030. Edgar Dunn forecasts $1.7 trillion in agentic commerce by 2030. J.P. Morgan estimates AI agents will handle 15–25% of all U.S. e-commerce by the same date.
According to a Checkout.com study of 4,000+ consumers, the average shopper is willing to spend $223 per agent-enabled purchase. But 85% of financial institutions surveyed believe current systems cannot handle high-volume autonomous agent transactions, per Accenture research.
The immediate economic question for Tempo is fee capture. At sub-$0.001 per transaction, Tempo's revenue model depends on extreme volume. If the chain processes 1 billion transactions per month at $0.001 each — roughly 385 TPS sustained — that yields $12 million in annual fee revenue. Stripe's business model likely relies less on protocol fees and more on the adjacent services (Bridge issuance, Stripe processing, Open Issuance platform fees) that Tempo's settlement layer enables.
This mirrors the foundational finding in blockchain economic-value research: fee revenue alone rarely sustains infrastructure. Value accrues to the entities that control adjacent services — issuance, compliance, onboarding, and distribution. Stripe's vertically integrated stack is explicitly designed to capture value at each layer.
Tempo's launch marks the first time a payments company of Stripe's scale has deployed its own blockchain to production with institutional partners actively building on it. The technical specifications — no native token, stablecoin-denominated fees, sub-second finality, EVM compatibility — are optimized for a specific thesis: that stablecoin settlement will displace a meaningful share of the $120 billion cross-border payments fee pool within the next five years.
Whether that thesis proves correct depends on adoption velocity. The 100+ launch partners and the involvement of Visa, Mastercard, and multiple banks provide distribution that most new chains lack. But Tempo faces competition from x402 (which already has 15 million transactions and Stripe itself as a user) and Visa's own card-native approach.
The three-way race underscores a structural shift. The question is no longer whether machine payments will exist. It is which settlement layer — stablecoin-on-chain, HTTP-native stablecoin, or extended card rails — will capture the bulk of the flow. Stripe, by building Tempo while simultaneously integrating x402, appears to be hedging across multiple outcomes. The data available today does not indicate which protocol will dominate. What it does indicate is that the infrastructure build-out is now fully underway.