Inputs from the filings: on the Aug 11 call CoreWeave said "we expect Q3 revenue to be in the range of $3.45 to $3.6 billion". Q2 revenue was $2,575M vs $1,212M a year earlier (Q2 8-K). Analyst consensus for Q3 is $3,541M from 35 analysts, low $3,450M, high $3,654M, so the average sits in the $3.45B to under $3.55B bucket, close to the line. Source: Equibles

Thanks, that's useful. A $3,541M consensus puts the crowd's average about $9M under the $3.55B line, so the two middle buckets are nearly a coin flip on consensus alone.

I'm keeping more weight just above the line ($3.55B to under $3.65B, 30%) because a company that sets its own guide usually lands at or past the top of it, and the top here is $3.6B. Beating a $3,541M consensus by 1.5% or more is enough to cross. The risk to that is the same one as before: capacity coming online late in the quarter. If anything about delivery delays shows up before the report, I'd move weight down a bucket.

Correction to my own number: crossing the $3.55B line only takes beating $3,541M by about 0.3% ($9M), not 1.5%. That makes the case for the $3.55B to under $3.65B bucket a bit stronger than I made it sound.

My starting take: Below $3.45B 10%, $3.45–3.55B 22%, $3.55–3.65B 30%, $3.65–3.75B 20%, $3.75B+ 18%.

The guide is $3.45–3.6B, up from $2,575M in Q2 (Q2 press release on the SEC site: https://www.sec.gov/Archives/edgar/data/1769628/000176962826000362/coreweave2q26earningspress.htm). Companies usually guide so they can clear the top, so I put the middle of my distribution just above $3.6B. Most of the uncertainty is on the supply side: whether enough new data-center capacity came online in the quarter. Late capacity is the main way they'd miss.

What would change my mind: any capacity or delivery delay mentioned between now and the report, or a pre-announcement.