.svg "logo"){kind=small}
On the 18th of September 2026 the AO500 closed at $6.30 per ton. Two and a half years earlier, in the first week of January 2024, it sat at $3.21. Over that period it bottomed at $3.16 and peaked at $7.07 in July 2026.
Those four numbers describe the carbon market's price history in a single line, and the ability to state it that plainly is fairly new. For most of the market's existence there was no agreed answer to what a credit costs, because every available source gave a partial view. Registries publish what sellers are asking, which is often well above what buyers pay. Exchanges publish settlement prices for a narrow, standardized set of contracts. Brokers know what deals actually close at, but those numbers stay private. A cookstove credit and a direct air capture credit both represent one ton of CO2 and can trade more than a hundred dollars apart, so a simple average across everything on offer tells you very little.
The AO500 was built to close that gap.
The AO500 is a weighted average price across the 500 most-retired carbon credit projects
The AO500 is AlliedOffsets' carbon price index. It tracks a weighted average price across the 500 carbon credit projects with the highest retirement volumes globally, weighted by the number of credits each project has retired, and it updates three times a week.
The word doing the most work there is retirement. A credit is retired when a buyer permanently takes it off the market to claim the emissions reduction: the moment the credit is actually used rather than listed, held, or traded. Building the index around retirement activity narrows the question to something far more useful than an average of everything on offer. Instead of asking what carbon credits are priced at in theory, the AO500 asks what they cost based on what the market is really buying.
How the index is built
Two decisions define the AO500: which projects go into it, and where the prices come from.
The project list is the top 500 by retirement volume. That single choice keeps the index tied to credits that are actually moving. Registries hold tens of thousands of projects, many of which issue credits that sit unsold for years, and including them would drag the benchmark toward prices nobody is paying. Restricting the universe to the most-retired projects means the AO500 reflects live demand.
Each project's weight is then set by how many credits it has retired, which is what separates the AO500 from a simple average. The difference is significant in a market this concentrated: a large cookstove or REDD+ project can retire millions of credits a year while a small project retires a few thousand. Under a simple average both would move the index equally, and a handful of small, expensive projects could lift the benchmark well above what the market is really paying. Weighting by volume means the biggest sources of retired credits carry the most influence, in proportion to the tonnage they supply.
The prices come from across the market rather than any single venue. AlliedOffsets collects over 17,000 price inputs a year from partners including brokers, traders, project developers, buyers, and marketplaces, then runs them through a pricing model that updates continuously as new data arrives. The model is built to screen out outliers and attempts at manipulation, so one unusual trade cannot move the benchmark. Breadth of sourcing is what keeps the index representative; the modeling is what keeps it stable enough to benchmark against week to week.
The AO500 and the Tradable20 have pulled apart
The AO500 has a companion index. The Tradable20 tracks the 20 most liquid and actively traded credits in the market, and watching the two move against each other says a lot about where the carbon market has gone since 2024.
They started in the same place. In the first week of January 2024 the AO500 sat at $3.21 and the Tradable20 at $3.28. Since then they have gone in opposite directions: the AO500 has climbed from a $3.79 average in 2024 to $6.04 so far in 2026, while the Tradable20 fell through 2025, bottoming at $1.94 in April, and has only partly recovered. The gap between them has averaged $2.58 per ton across 2026 and reached $4.12 at its widest in June.
That gap is why the choice of benchmark matters. A corporate checking a supplier quote against the Tradable20 when its portfolio actually looks like the AO500 is measuring against the wrong market, and at today's spread the error runs to several dollars a ton.
The distinction is straightforward. The AO500 tells you what a credit costs across the market as a whole, based on what buyers are genuinely retiring. The Tradable20 tells you what the most liquid, standardized credits are trading at. Most buyers need the first; desks working in the traded end of the market need the second.
What the AO500 is not
Knowing what the index deliberately leaves out is as useful as knowing what goes into it.
It is not a registry listing price, because listings show intent rather than outcome. It is not an exchange settlement price, which is valuable for trading but covers too narrow a slice of contracts to represent the wider market. It is not a quality rating: the AO500 measures price, not integrity. Buyers who want both lenses at once can look at the Quality-Weighted Carbon Price Index that AlliedOffsets built with Sylvera, which pairs pricing data with project quality assessments.
Independence underpins all of it. AlliedOffsets does not broker, trade, or issue credits, so the index has no position to talk up or down. That neutrality is a large part of why it travels beyond our own platform: analysts at CRU Group, among others, have used the AO500 to describe carbon credit price movements in their own market commentary.
See the current AO500 price
Every figure in this post comes from the AO500 and Tradable20 price indices, both of which are archived in full and updated three times a week.
The live AO500 price, a side-by-side chart against the Tradable20, and the methodology behind both indices are all on our AO500 index page.
Explore the AO500 index here.