Render Opens RNP-024 Vote on Forth AI Gateway

Render Network is preparing to open the first community voting round on RNP-024, a proposal that would bring Forth into the ecosystem as a privacy-focused gateway for AI models and distributed compute. According to the official Render Network announcement, voting is scheduled to begin October 27 at 5 p.m. PT and remain open through October 30, with a snapshot taken shortly before voting begins to determine eligible RENDER voting power.

The underlying RNP-024 specification proposes Forth as a single API layer capable of routing users, applications and autonomous agents toward different AI models and compute environments. The proposal would extend Render’s AI strategy from supplying GPU capacity toward aggregating demand and deciding where compatible workloads should execute. The RNP remains in Draft status in Render’s official repository.

First round of voting begins Tuesday Oct 27th, 5pm Pacific time, for RNP-024, which introduces Forth to the Render Network ecosystem as a gateway to privately access AI models. A snapshot will be taken shortly before.https://t.co/epTBb1Qo3C

Review the details, including the…

— The Render Network (@rendernetwork) October 26, 2026

Forth Would Route AI Across Multiple Privacy Levels

Forth is designed to abstract authentication, model selection, metering, payment and infrastructure routing behind one interface. Users could choose between workloads running on private hardware, anonymized cloud access and compatible distributed compute, while Forth could automatically select an appropriate route when a specific model or environment is not requested.

That privacy model is not uniform across every execution path. Workloads running on hardware controlled by a user or organization can remain inside that private environment, while external cloud requests may still reach third-party providers with the user’s identity separated from the request. Future trusted execution environments and cryptographically verifiable confidential compute are planned capabilities rather than features RNP-024 establishes as currently deployed. The distinction resembles NEAR AI Cloud’s confidential-inference architecture, where privacy guarantees depend on the actual infrastructure handling a workload.

The proposal builds on Render’s earlier Forth AI gateway design, which introduced the project as both a model-access interface and a potential source of demand for Render Network compute. Forth would not require every AI request to run on Render hardware, because workloads could also be routed toward external model providers when that route is technically or economically appropriate.

RENDER Incentives Target Paid AI Usage

RNP-024 also proposes a capped incentive program of up to 500,000 RENDER funded from existing Network growth and operations resources. The program would reward economically measurable usage rather than account creation or free requests, with paid inference, media generation, agent API usage and other qualifying services contributing to a user’s reward score.

Payments are designed to include RENDER and USDC through x402 on Solana, alongside conventional card funding through MoonPay Commerce. RENDER-funded usage would initially receive a 1.10x scoring multiplier, while eligible workloads routed through Render Network compute would receive a proposed 1.25x multiplier. The payment design creates preferential economics for RENDER without making the token mandatory for accessing Forth. Similar machine-payment infrastructure is already appearing through x402-based autonomous services on other networks.

The reward structure also limits how quickly incentives can be distributed. Monthly rewards would initially be capped at 50,000 RENDER and at the RENDER equivalent of 20% of qualifying monthly paid usage, with the full program limited to 500,000 RENDER over no more than 12 months. Those limits are designed to tie token distributions to actual economic consumption rather than subsidizing artificial request volume.

Render’s direction reflects a broader shift among decentralized GPU networks toward packaged AI services rather than raw compute marketplaces alone. io.net, for example, recently added Reflection Beam 1.0 to its inference platform, exposing a large open-weight model through distributed GPU infrastructure. Forth differs by positioning itself primarily as a routing and demand-aggregation layer capable of selecting among multiple infrastructure types.

For now, RNP-024 remains a governance proposal rather than a completed Forth deployment into the Render Network economy. The first vote will determine whether the proposal advances, while production readiness, payment integration, anti-Sybil controls and deeper Render Compute routing remain later implementation stages if the governance process approves them.