As of our 2026-09-22 pull, our model grades Ethereum F / Failed at a score of 77.9.

Two things in our own dated record explain why that sentence is stranger than it looks, and neither requires you to take our word for anything.

A CryptoGrade letter grade is not our score rounded into a bucket. Ethereum is the cleanest illustration of that we have ever published, because it is one asset, one week, and two moves in the same direction that did opposite things to the letter.

Every figure below is quoted from our own record as it returned on 2026-09-22. Nothing is derived, averaged or reconstructed.

Unranked is not the same as last. An asset without a rank has not been sorted to the bottom of our book. It is outside the ranked list entirely, which is a different statement, and at 77.9 Ethereum sits nowhere near the bottom of our scores.

Our own one line read on the asset, quoted word for word from our record:

That line and the letter F sit in the same record, on the same day, about the same asset. We are not going to pretend that is intuitive. It is, however, exactly what our framework is built to do, and the rest of this piece is the explanation.

Our record for Ethereum carries 94 dated readings, from 2026-06-20 to 2026-09-22. Here is every point at which the grade or the score changed across that whole span.

87 of those 94 readings are grade A. Every reading from 2026-06-20 through 2026-09-15 is A, with the score moving inside a narrow band between 84.4 and 89.6 the whole time. Nine score changes across roughly three months, and not one of them touched the letter.

Then the letter changed, on the smallest move of the three that followed.

Put the last two moves side by side.

2026-09-15 to 2026-09-16. Score 88.7 to 87.1, a fall of 1.6. Grade A to F.

2026-09-18 to 2026-09-19. Score 87.1 to 77.9, a fall of 9.2. Grade F to F.

The larger move did nothing to the letter. The smaller one moved it the entire length of our scale.

If the grade were a bucket you drop the score into, that is impossible. A 9.2 point fall from 87.1 would have to cross more boundaries than a 1.6 point fall from 88.7, not fewer. So the grade is not a bucket, and our own record is the thing that says so rather than us asserting it.

What we are deliberately not doing here is telling you what caused either move. That distinction matters enough to have its own section below.

Our research record carries a standing assessment of Ethereum across the six dimensions our framework scores. Summarized, in our own record's terms, as of the 2026-09-22 pull:

Tokenomics and Value Accrual. ETH has no maximum supply and a broad holder base. Our record credits fee burn and staking economics as supporting value accrual, and holds the dimension mid-range because circulating supply is large and the evidence in front of it does not quantify current net issuance or dilution risk.

On-chain Usage and Economics. This is the strongest dimension in our record. It cites total value locked of about $54.1 billion and 30 day DeFi fees of about $318.4 million, both in the top band, alongside healthy development activity on the go-ethereum client.

Security, Decentralization and Durability. Our record describes a mature validator and security profile and a long operating history. It stops short of a perfect read because the evidence in front of it does not include audit detail, node or client diversity, or a full cycle failure test.

Ecosystem and Moat. Our record puts Ethereum at the top of this dimension: the reference smart contract ecosystem, the deepest DeFi base, strong developer activity, and high switching costs relative to credible rivals.

Relative Valuation. Market capitalization and fully diluted valuation are both about $333.2 billion in our record, which it reads as no obvious overhang from future dilution. It calls the dimension roughly neutral, because it has no peer relative fees-to-market-cap or network-value-to-transactions context to judge against.

Market Structure. Our record reads liquidity and tradability as good on the strength of size. It holds back from a maximal read because it has no spread, depth or realized volatility data.

Our standing thesis, summarized: the evidence supports Ethereum as a high quality, high usage asset with strong network effects, deep DeFi activity and sustained client development, with tokenomics supportive but not enough to justify a maximal score absent clearer dilution and accrual data.

The full published assessment and its supporting research are behind a free account. Our own record says so in the line it returns to logged out readers: "Sign in to read the latest published assessment and its supporting research."

Read those six dimensions again and count the qualifications. Four of the six say, in our record's own words, that something it would want is absent:

We publish those rather than smoothing them over, and we think that is the more useful document. A research note on the second largest asset in crypto that has no gaps in it is not a more thorough note. It is a less honest one. Where our evidence stops, our record says where it stopped, and the score reflects the evidence we actually had rather than the evidence we wish we had.

This is also why the Security dimension reads as strong rather than perfect for an asset with Ethereum's operating history. Our framework does not award a dimension for reputation.

The one sentence to take from Ethereum's record: at CryptoGrade, a score and a letter are two different statements about an asset, and you can read both on our site.

The score is an absolute measurement against our six weighted dimensions. The letter carries the score and the pass or fail screens together. That is why 77.9 and F can be true of the same asset on the same day, and why our score for Ethereum falling 9.2 points did not change the letter, while a 1.6 point move did.

You can read our dimensions, weights, screens and bands for yourself at https://cryptograde.ai/framework .

Our record on Ethereum changed letter on 2026-09-16. Nothing on our website went and told anybody.

A reader who checked our Ethereum page on the 15th and again on the 22nd saw 88.7 and A, then 77.9 and F, and never saw that the letter turned over on the 16th while the score was still 87.1, or that the big score move three days later did nothing to it. The whole of what this piece is about happened in between two visits.

That gap is what Alerts closes. Alerts is free, it needs a free CryptoGrade account, and it sits on every token detail page including Ethereum's, so you can turn it on for the assets you actually care about and leave the rest alone.

A free account gets you two concrete things this piece has referred to throughout: Alerts on the assets you pick, and the full published assessment behind the grade rather than the summary above.

Create one at https://cryptograde.ai