Solana Foundation Maps AI Agent Payment Plan at Breakpoint 2026
The proposal combines financial benchmarks and pay-per-request AI access with a card-to-stablecoin route to fund agents.

Rishin Sharma, head of AI initiatives at the Solana Foundation, outlined plans at Breakpoint 2026 in London on Nov. 17 to support AI agents with payments and paid model access.
The roadmap combines financial benchmarks for AI models, model inference access and a card-to-stablecoin onramp designed to fund agents. Inference is the step in which a trained model generates an answer, and the plan would let agents pay for that service.
Sharma pointed to x402, a standard for machine-to-machine payments. He also highlighted Solana payment channels, which he said can handle up to one million payments per second, and Pay.sh, a service that provides paid access to more than 100 APIs.
The presentation cited partnerships with Google, Alibaba and ByteDance for pay-per-request access to AI models. Under that setup, an agent pays for each request instead of using a monthly subscription.
Solana.ai, a hub for AI projects on the network, launched a week before Sharma’s Nov. 17 talk and listed more than 120 teams. Payments were the leading use case among those teams.
Breakpoint 2026 runs Nov. 15-17 at Olympia London, following Solana’s earlier announcement of the conference.