2026-01-18 09:34:14

AI startups find it easy to raise funds, but achieving scale and profitability is becoming increasingly difficult—this is a real problem.

Saying that AI is more quickly monetized than during the internet bubble era is only half true. Monetization does not equal profit—the key difference lies in the cost structure.

Once traditional software is developed, the marginal cost of replication approaches zero, which is why it can maintain a 90% gross profit margin. AI is different; each user query and each model inference requires spending money on computing power, and this expense cannot be avoided.

What is the result? Even the most outstanding AI startups have their gross margins tightly squeezed to 50%-60%. This figure looks decent, but compared to the funding expectations of high-growth startups, it’s quite disheartening. In the long run, this cost structure limits the imagination for scaling up.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

4 Likes

Reward
4
6
Repost
Share

Comment

0/400

MEVHunterNoLoss

· 7h ago

The cost of computing power is indeed the Achilles' heel of AI. Rapid fundraising but difficulty in making money is a major flaw.

View OriginalReply0

LiquiditySurfer

· 7h ago

Raising funds is easy, but cashing out is hard. This trick has been tired of seeing long ago. The key is that old saying—without a good cost structure, no matter how much money you have, it's useless. --- 50%-60% gross profit sounds okay? If you look at the historical accounts of internet software, it seems quite modest. --- Burning money on computing power is unavoidable. Frankly, the ceiling for AI startups is written in their genes. --- Cash out ≠ making money. This distinction is very sharp and hits the pain points of many projects. --- Instead of celebrating fundraising, it's better to think about how to understand the cost structure. Otherwise, you're just working for cloud providers. --- The fact that gross profit margin is being squeezed actually reflects a structural problem—economies of scale are not achieved, which is a hard flaw.

View OriginalReply0

MindsetExpander

· 7h ago

Raising funds quickly doesn't necessarily mean lasting long, and that's truly ruthless. The cost of computing power is indeed the Achilles' heel of AI startups, no wonder everyone is competing over efficiency.

View OriginalReply0

NeverPresent

· 7h ago

This is what I've always wanted to say: easy financing = a viable business model, two different things. Gross profit being stuck at 50-60%, the wall of computing power is really insurmountable.

View OriginalReply0

GateUser-2fce706c

· 7h ago

I've already said this before, the secret to AI wealth this time isn't in the application layer, but in computing power and model optimization. With gross profit margins squeezed to 50-60%, the financing story can't be spun convincingly, and capital will wake up sooner or later.

View OriginalReply0

ImaginaryWhale