Meta is up 30% this month on Muse hype. I own it, and I think the market is still pricing it as just an ad company.

I'm long META, bought it for the ad machine plus the AI optionality, and this month the optionality finally showed up. The stock closed Thursday at $779, September is on track to be its best month since July 2013. Muse, the personal agent launched on September 8, has overtaken ChatGPT as the top free app in the US and Canada, and price targets have been raised across the Street. JPMorgan went to $920 and wrote that Muse could become the most widely used consumer AI app since ChatGPT.

Bear case is still active, Q2 capex was $31.1B, and free cash flow fell to $784M on $60.8B of revenue. Full year capex guidance is $130-145B. When Q2 printed, the stock sold off 8-10% after hours on a soft Q3 revenue guide and an EPS miss. Of 49 EPS revisions for 2026, 45 have gone down, because depreciation from all that spending is starting to hit earnings.

Most importantly, Meta hasn't published a single official Muse usage or revenue number. Some download figures circulating this week are unconfirmed. On top of that, OpenAI, Google and Amazon are all shipping agents of their own. A 30% move in a month on app store rankings and analyst notes is exactly what a top looks like.

I feel people keep valuing Meta as if Muse is the whole bet, the core business in Q2 was ad revenue of $59.4B, up 27%, with impressions up 14% and price per ad up 12%. Both volume and pricing are growing at the same time, at this scale. Even after the rally, the stock sits around 22x forward earnings. To me that's a fair multiple for the ad business alone, which means Muse is being priced close to zero. Muse is also already monetising with paid tiers at $20 and $100 a month, and TD Cowen models Muse revenue going from $55M this year to $1.8B in 2027. You don't need to believe the $27B 2031 number to see that the option is worth more than nothing.

The capex is also paying off in the ad engine before Muse contributes anything. On the Q2 call, Zuckerberg made the point that selling intelligence carries far better margins than renting out compute.

Meta reports Q3 in late October, if management discloses Muse usage figures and holds capex inside the $130-145B range, I think the stock holds above its pre Connect level through the end of the year. If Muse gets no hard numbers on the call, or capex guidance goes up again, then this month was a sentiment rally and I'm wrong on the timing.

For the bears here, what's the actual case against Meta at 22x forward if the ad business keeps growing in the high 20s? And for everyone, what Muse number on the October call would change your mind either way?