Render (RENDER) token overview: a record month of paid jobs, the burn-mint math, and the Forth vote

Render Network had its biggest month of 2026 in October. By DefiLlama’s count, customers paid $834,203 for render jobs, up from $232,323 in September. That is about 3.6 times the prior month, and more than the first six months of 2026 combined ($818,237) (DefiLlama, Render Network BME).

In the same week those numbers came out, RENDER holders are voting on RNP-024, a proposal to launch Forth, an AI gateway meant to route paid AI demand toward the network. The final vote closes November 9 at 5 PM PT (@rendernetwork).

This overview covers how the token works, what October’s spike shows and what it doesn’t, and what Forth would and wouldn’t change.

What Render is, in one paragraph

Render is a decentralized GPU network. Artists, studios, and more recently AI customers pay for GPU work, and node operators with idle GPUs do the work and earn RENDER. It began as a marketplace for 3D rendering, built around OTOY’s Octane renderer, and has spread into general and AI compute. RNP-019 created a separate compute subnet for AI and general-purpose workloads, and RNP-021 extended it to enterprise-grade compute (RNP-022). The token moved from Ethereum (RNDR) to Solana (RENDER). By October 2025, 423M tokens had been upgraded, about 77% of supply, per the Foundation’s own Year 3 budget proposal (RNP-022).

How the token works: burn-and-mint equilibrium

Render runs on a burn-and-mint equilibrium (BME), adopted through RNP-001. It has two sides that work independently:

• Burn (demand side). Each job is priced in USD. The customer pays, and RENDER worth that USD amount is burned. Under DefiLlama’s methodology, the on-chain burn is the 95% node-operator share of gross job spend. The other 5% is a Network Operator service fee paid to OTOY off-chain (DefiLlama methodology).
• Mint (supply side). Separately from how much work is done, the network mints new RENDER on a schedule. Those tokens go to node operators, grants, and Foundation operations.

Because jobs are priced in dollars, a falling token price means more RENDER gets burned for the same job. That is the stabilizer built into the design (RNP-001). Whether net supply grows or shrinks in a given month depends on one comparison: dollars of paid work against tokens minted.

The supply side in numbers

RNP-022 sets Year 3 BME emissions at 5.9M RENDER, running from Dec 20, 2025 to Dec 19, 2026. The split:

Render node rewards continue at 15K RENDER per week for job rewards plus 2.8K per week for availability, or 17.8K per epoch. RNP-022 puts that at roughly 71K–90K RENDER a month.

Spread evenly, 5.9M a year averages about 492K RENDER a month in emissions. That’s our arithmetic. Actual monthly distribution varies.

The demand side: October in context

Monthly gross job spend, per DefiLlama:

On the token side, MSB Intel reported that by October 22 Render had already burned 243,193 RENDER for the month. That made October its biggest month on record, more than double the previous high of 120,928 RENDER in September 2025 (MSB Intel). For the full month, an independent burn tracker that is not an official Render account counted 363,662 RENDER burned (@NEWRenderBurn).

Set that against the ~492K monthly average from RNP-022 and October’s burn covered roughly three-quarters of average monthly emissions. That’s a big change from earlier in the year. It also means BME was still net inflationary in its best month on record.

A note on sources: trackers disagree on October’s dollar figure. The independent burn tracker valued October’s burns at $667,615. Some posts on X cited about $697K. DefiLlama’s methodology grosses up the on-chain burn to include OTOY’s 5% fee. We use DefiLlama’s $834,203 here because its method is published, and we treat every figure as an estimate built from on-chain burns, not audited revenue.

What the October spike does and doesn’t show

What it shows. Paying customers pushed far more work through the network than in any month this year. Render’s own case studies point to the kind of work involved. STURDY rendered visuals for Rauw Alejandro’s Latin Grammys set on Render. Final renders that would have taken 10 days locally finished in under 24 hours, and one 1,471-frame sequence went from 12 hours on a workstation to 40–50 minutes on the network (@rendernetwork). Immersive pieces for the Render x Solana “Beyond the Frame” show at FRAMELESS London on Nov 16 were also rendered on the network (@rendernetwork).

What it doesn’t show (yet). That the demand is steady. DefiLlama’s daily series shows October’s spend was concentrated. Five days, Oct 16–18 and Oct 26–27, account for about $518K, or roughly 62% of the month. The pace since then has been far slower. DefiLlama shows $49,787 in fees over the trailing seven days and about $75K for November to date as of this morning (DefiLlama).

Lumpy demand isn’t a bad sign in itself. Rendering is project-driven, and large productions arrive in batches. But one big month is a data point, not a trend. The test is whether the baseline between spikes keeps rising.

Two other caveats:

1. Job spend is not profit. Most of the payment pays for GPU work. The burn is the token-level effect, not a margin.
2. The AI compute side is still small in the public data. MSB Intel reported that September node rewards on Dispersed, Render’s AI compute subnet, more than doubled to 3,784 RENDER (MSB Intel). That is growing, but it is small next to the render node budget of about 15K RENDER per week.

RNP-024: what Forth is and what it would change

Forth is the proposal aimed squarely at that second caveat. According to the RNP-024 text, authored by Render Labs:

• What it is: a “privacy-first AI gateway, routing, and payment layer.” Users, apps, and autonomous agents get a single API endpoint that can route to user-controlled hardware, anonymized third-party and frontier models, Render Network and other distributed GPU compute, and, later, confidential-compute infrastructure.
• Who builds it: OTOY leads development. The RNP says operational responsibility for the live service will be disclosed before launch.
• How it gets paid: wallet-native payment via x402 on Solana in RENDER or USDC, plus card payments through MoonPay Commerce. Chat usage draws on prepaid credit. The Foundation also intends a product-level discount for paying in RENDER, with about 5% given as an example parameter.
• The incentive budget: up to 500,000 RENDER for a performance-gated incentive program, plus a separate 150,000 RENDER marketing budget. That’s a maximum Foundation exposure of 650,000 RENDER, funded from existing Foundation resources. The RNP states it does not raise the emissions cap.

How the incentives are fenced in

The design rewards spending, not sign-ups:

• Users earn non-transferable Forth Points based on net paid usage. Free tiers, promo credits, refunds, and circular activity don’t count.
• Multipliers: 1.10x for paying in RENDER, 1.25x for work routed to eligible Render compute, and up to 1.30x for 12+ consecutive paid weeks. Combined multipliers are capped at 1.50x.
• Only the first $1,000 of monthly paid usage per participant earns the full score.
• The monthly reward pool is the lesser of 50,000 RENDER, the remaining allocation, or the RENDER equivalent of 20% of qualifying paid usage. In a slow month, little gets distributed.
• Rewards unlock in tranches: 100K pilot, then 150K, then 250K, and the Foundation can decline to unlock later phases.
• The Foundation must publish reporting on paid usage, the share paid in RENDER, Render-routed compute, incentives distributed, and related burns.

The honest limitation

The RNP itself flags the main risk. Early Forth usage may run on centralized third-party models, and a payment for a third-party model can’t simply be burned because the model provider has to be paid. For that traffic, the Foundation commits only to developing a methodology later, with options such as a share of router revenue, RENDER-denominated fees, or periodic buy-and-burn. So Forth’s effect on BME depends on how much of its traffic actually reaches Render compute, which is why that route earns the higher multiplier.

Where the vote stands

Round one closed with 100% Yes (@rendernetwork). The Chain Observer reported about 2.4M RENDER voting in favor with no opposing votes (The Chain Observer). The final vote opened Nov 3 at 5 PM PT. It needs at least 50% approval and a 15% quorum (@rendernetwork), and it closes Nov 9 at 5 PM PT. We couldn’t read a live tally from the Nation voting page, so we don’t report one here. In round one, the bar that mattered was turnout, not approval.

How to read Render from here

The simplest way to track Render is one ratio: dollars of paid jobs per month against tokens minted per month. October moved that ratio further than any month this year. Forth is the network’s attempt to add a second, AI-driven demand source that doesn’t depend on a few large render projects.

If you’re comparing Render with other compute networks, the same discipline applies as in our pieces on Akash provider economics and cost per accepted result. Look at what customers actually pay, how steady that payment is, and how much of it reaches the token.

What to watch

• RNP-024 final result (closes Nov 9, 5 PM PT). Did it clear the 15% quorum, and what was the final Yes share?
• November job spend. Does the baseline between big projects hold above summer levels (~$85K–$165K a month), or was October a one-off?
• Burns vs. emissions. Whether monthly burns get closer to the ~492K RENDER monthly emissions average, and what the Year 4 allocation RNP proposes when Year 3 ends Dec 19.
• Forth’s first public report. In particular, the share of Forth usage paid in RENDER and the share routed to Render compute rather than third-party models.
• AI compute subnet rewards. Whether Dispersed grows from a few thousand RENDER a month toward the scale of render node rewards.
• Breakpoint week (Nov 16, London). Render’s showcase at FRAMELESS. Watch for new studio or AI customers, not just exhibitions.

Research and commentary only, not financial advice. Figures are from the linked sources as of Nov 7, 2026 and may be revised.