PayAI Network has launched x402 batch settlement on Solana in public preview, introducing a payment-channel model designed for AI agents, APIs and other services that generate large numbers of small usage-based charges. According to the Solana Foundation’s official Payment Channels announcement, the architecture lets a user fund a channel once, authorize subsequent usage offchain and settle accumulated spending on Solana later, avoiding a separate blockchain transaction for every service call.
PayAI’s implementation applies that architecture specifically to x402, the open HTTP payment protocol now governed by the Linux Foundation’s x402 Foundation. The public preview currently targets USDC micropayments on Solana and is designed for workloads such as AI inference, metered APIs, data retrieval and agent tools, where the value of an individual request can be comparable to or lower than the cost of settling it independently.
Payment Channels Move Each Request Offchain
Under the Solana implementation, the payer first deposits funds into an onchain escrow channel. Individual requests then update a cumulative payment commitment through signed Ed25519 vouchers rather than Solana transactions. The merchant or facilitator can subsequently redeem the latest authorized amount onchain and distribute the settled funds. The service usage is metered offchain, while custody and eventual financial settlement remain anchored to the Solana payment-channel program.
The x402 Foundation formalized the mechanism in its shared specification after the Solana batch-settlement implementation was merged into the repository on September 25. The standard defines batch settlement as a flow where access can be granted after a valid cryptographic commitment even though value moves later. That separates payment authorization from immediate blockchain settlement, unlike conventional x402 exact flows that settle each request individually.
All compatible Solana implementations use the canonical payment-channel program CHNLxYvVA28MJP9PrFuDXccuoGXAx7jBacfLEkahyGsX, maintained within the Solana Foundation ecosystem. That common program allows other x402 facilitators to build against the same channel infrastructure rather than creating incompatible settlement contracts. The interoperability comes from a shared program and specification, not from PayAI controlling the underlying payment primitive.
The model extends a broader shift toward machine-native payments. Injective has already deployed x402-based AI agent payments on mainnet, while Coinbase Agentic Wallets combine autonomous wallets with x402 infrastructure. Batch settlement addresses a different bottleneck: reducing how often those machine payments must actually consume blockchain settlement resources.
PayAI Estimates Major Savings for BlockRun
PayAI uses BlockRunAI as its main example of the potential cost reduction. The company says BlockRun settled approximately 18.5 million payments over a 30-day period using individual x402 transactions, generating roughly $18,400 in network fees. PayAI estimates that processing an equivalent volume through batch settlement could reduce blockchain fees to less than $1. The comparison is illustrative rather than a measured before-and-after production test.
The economics follow directly from amortizing settlement across many requests. A million $0.001 service calls can accumulate as signed commitments against one funded channel instead of requiring one million separate blockchain settlements. That structure is particularly relevant when agents continuously purchase small amounts of compute, data or model output. Batching reduces settlement overhead without eliminating the underlying cost of the service, facilitator operations or channel management.
PayAI’s current preview also has operational limits. Its rollout uses bounded channel capacity, and batch-settlement requests require merchant authentication. PayAI’s documentation further warns that a successful claim is not necessarily the final merchant payout because distribution must complete and the receiving USDC balance should be reconciled. The release therefore establishes working infrastructure, but not yet evidence of unrestricted production throughput or broad merchant adoption.
The development complements other approaches to lowering payment friction, including Sui’s protocol-level gas-free stablecoin transfers. PayAI takes a different route by leaving individual usage events offchain until settlement is economically worthwhile. Its significance will ultimately depend on whether merchants begin processing sustained commercial workloads through these channels, rather than on the theoretical number of micropayments the architecture can aggregate.