Meta finally saw a long-awaited surge, with the pre-market share price back above 650+. After reviewing everything, it feels like Meta’s current situation is somewhat similar to last year’s Google playbook.

Both went through a prolonged period of decline. Last year, Google was punished by the market for the narrative that AI large models would upend traditional search, with the stock price being continuously hit.
This year, Meta has faced even more nonstop fud. Llama open-source model technology lagged, the flagship model was delayed, the AI team was reorganized, and all of that came on top of continuously rising capex. The market has started to seriously suspect that Meta is using the cash flow from its advertising business to subsidize an AI arms race with no visible endpoint.
Meta’s stock price fell from its all-time high to its low point, dropping as much as 35% at one stage—extremely pessimistic.
However, both companies have since experienced a fundamental turnaround.
Last year, Google proved itself with several consecutive quarters of earnings reports that beat expectations, along with its most cutting-edge Gemini large model. AI would not only fail to erode its existing business, but would instead become a new growth engine. After the reversal, Google became one of the few global tech giants with an end-to-end AI ecosystem.
And recently, Meta is doing something similar. It has launched Muse Spark 1.1, its flagship large model, which is said to have already surpassed the Gemini model on multiple benchmarks.
Most importantly, Meta is charging externally for the first time via a large model API, with pricing at only one quarter of OpenAI’s and Anthropic’s.
In addition, at the application layer, Meta will sell AI agents to enterprises; at the compute layer, Meta is also planning a cloud business, building AI cloud infrastructure and selling compute power externally. This truly opens up a new source of AI revenue, demonstrating to Wall Street that its AI capital expenditures make sense.
From a short-term stock price perspective, the key thing to watch next is whether Meta can break through the resistance zone of 650–670. Personally, I still think there’s plenty of room for catch-up gains over the coming period, especially since the 18–19x forward PE is already the lowest level since the pandemic.