Render's official October report: burns nearly tripled, frames fell, and Salad started daily test burns

On November 9 the Render Network Foundation published its October monthly report. The official numbers confirm the shape of the month we described in our Nov 7 overview, and they add a detail that the dollar figures alone bury: frames rendered went down.

The Foundation’s table puts October burns at 363,662 RENDER across all subnets, up 159% from 140,532 in September, against 492,132 RENDER of emissions, unchanged month over month. In dollar terms, it records $677,300 of RENDER burned (USDC equivalent) on the Render Network and $5,134 on Dispersed. Frames rendered fell 14%, from 1,733,953 in September to 1,486,238 in October (Foundation October report).

Separately, Salad — the compute marketplace approved as a Render subnet under RNP-023 — said that as of the week of November 6 it is burning RENDER on-chain every UTC day as a test, not as the finished Milestone 3 settlement system (Salad blog, Nov 6).

The official October table

From the Foundation report’s metrics images:

Node-operator rewards, also official:

The asterisk on Foundation-node rewards appears in the report table; the published images we read did not include a footnote explaining it.

October’s 363,662 RENDER burned covers about 74% of that month’s 492,132 RENDER emissions. That matches the “roughly three-quarters” comparison we made on Nov 7 from an independent burn count and an average of the Year 3 emissions budget. The Foundation has now made both sides official. BME was still net inflationary in the best burn month on this table: more tokens were minted than burned.

The October emissions figure, 492,132 RENDER, is also essentially the Year 3 budget from RNP-022 — 5.9M RENDER from Dec 20, 2025 to Dec 19, 2026 — spread evenly over a month. We had used ~492K as arithmetic. The report shows the network actually issued that amount in October.

How this sits next to DefiLlama

In the Nov 7 overview we used DefiLlama’s gross job-spend series, which put October at $834,203, up from $232,323 in September (DefiLlama, Render Network BME). DefiLlama’s method grosses the on-chain burn up to include OTOY’s 5% Network Operator fee.

The Foundation’s official dollar figure is lower: $677,300 on the Render Network, or $682,434 if you add Dispersed. September is much closer — the Foundation’s $226,853 (plus $2,230 on Dispersed) sits next to DefiLlama’s $232,323. We are not going to force those October numbers to reconcile. Different trackers are counting different things, and the official series is the one the Foundation is willing to put its name on. The direction of the month is not in dispute: dollar burns rose sharply; token burns rose sharply; emissions did not.

One independent figure from our earlier piece now has official confirmation. The unofficial burn tracker we cited for October had 363,662 RENDER burned. That is exactly the Foundation’s “total all subnets” number.

Frames fell. November has been quiet. The spike was lumpy.

A record burn month with fewer frames is the report’s most useful caution. Customers paid for more dollars of work, not for more frames. That can happen when jobs are heavier — higher resolution, more samples, a new Performance tier, or a handful of large projects. The report itself points at the last of those. It says the network has now passed 82 million frames rendered in total, and that the 3D works for BEYOND THE FRAME at FRAMELESS London on November 16 would have taken 4.7 years continuously on an RTX 5090, but were finished on the network in a couple of weeks (Foundation report).

It does not show that baseline demand has stepped up permanently. Two other official or public series still argue for lumpiness:

• Render Network job rewards to node operators fell from 75,000 RENDER in September to 59,970 in October, and availability rewards fell with them. Dispersed rewards rose, but only from 3,784 to 4,436 RENDER — still a small line next to the render subnet.
• DefiLlama’s daily series, retrieved this morning, still shows October concentrated in a few days, and November running much colder: about $61,938 from Nov 1 through Nov 10, and about $50,888 over the trailing seven days (DefiLlama). That is the same pattern we flagged on Nov 7, and it has not reversed.

The test remains whether the baseline between spikes holds above summer levels, not whether one month can post a record.

Salad’s daily burns are a test, not Milestone 3

RNP-023, approved earlier this year, is how Salad is supposed to join Render’s burn-and-mint model. The revenue splits in the RNP and in Salad’s later architecture post are:

(RNP-023; Salad Milestone 3 architecture)

What started this week is narrower than that design. Per Salad CEO Bob Miles (Nov 6 post):

• Each UTC day is a burn epoch for Salad Container Engine.
• When the day closes, Salad prices that day’s Chef payouts and usage revenue at the same 24-hour TWAP it uses for Chef rewards, then sends that amount of RENDER to the Render Network burn address on Solana.
• Render mints back the Chef payout portion; the rest stays burned.
• The process is automated, with a human hold if a day’s numbers look wrong.
• Salad is “starting incrementally and will step up from there.”
• SGS settles monthly; its first burn, covering November, is scheduled for December.
• This is not the full BME implementation. Milestone 3, still being built, is supposed to replace these test burns with escrowed customer deposits, metered usage settled on-chain, and Chef rewards paid from the Reward Pool.

We have not independently totaled the daily Salad burns since they started, so we do not report a running figure here. The right way to read the announcement is: a second, automated burn stream has begun, under test conditions, with an explicit warning not to treat it as finished BME settlement. If Milestone 3 works as designed, Salad’s burns and Chef rewards will be comparable in the same epoch, at the same price. That comparison does not start until the test is replaced by the real engine.

The Foundation report itself does not cover these November test burns. It is an October document. It does note that RNP-024 (Forth) passed its first vote in late October, that a second vote would follow in early November, and that Forth is in development with a launch still to be announced. The final Forth vote closed November 9 at 5 PM PT; we are not reporting a result here because we could not independently confirm the official tally and quorum from the public Nation page in a way we trust.

What to watch

• November job spend and official burns. Does the baseline hold above summer levels, or does October remain a spike with a quiet month after it?
• Burns vs. the 492,132 RENDER monthly emissions run rate. Official October coverage was about 74%. Salad’s test burns, if they scale, are an additional stream — once they can be totaled on-chain.
• Salad dashboard and first SGS burn. Daily SCE burn vs. mint figures, and the December burn that is supposed to cover November SGS usage.
• Milestone 3, not the test. Escrowed deposits, on-chain metered settlement, and a public stats API are the signs the temporary daily burns have been replaced.
• Dispersed. Official October rewards were still 4,436 RENDER. Forth is the proposal aimed at routing more AI demand that way.
• BEYOND THE FRAME (Nov 16, FRAMELESS London). Watch for new paying customers after the showcase, not just the exhibition itself.

Research and commentary only, not financial advice. Figures are from the linked sources as of Nov 10, 2026 (America/New_York) and may be revised.