DeepSeek Slashed Prices 75%. AI Costs Are Still Rising.

Key points

• DeepSeek cut the price of its V4-Pro large language model by 75% in mid-2025, but many business customers say their AI bills are still climbing.
• A single user request handled by an AI agent, software that carries out multi-step tasks on its own, can trigger 35,000 or more billable units of text, called tokens, where a basic chatbot would trigger a few hundred.
• One realistic enterprise query on a leading model costs between $0.10 and $0.40; at one million queries a month, that is a six-figure line item before any other expenses.
• Several enterprise software vendors are privately reporting negative gross margins on their heaviest AI users, meaning those customers cost more to serve than they pay.
• OpenAI recently offered every startup in the Y Combinator accelerator $2 million in free API credits, a sign of how expensive it now is to build an AI-native product from scratch.

DeepSeek, the Chinese AI lab, cut the cost of its V4-Pro model by 75% earlier this year. For companies building AI products, that should have felt like a pay rise. For many, it has barely moved the needle.

The reason sits in how modern AI systems are actually built.

A basic chatbot is simple: one question from the user, one answer from the model. An AI agent is different. It is software that breaks a single user request into many smaller steps, planning what to do, searching for information, using tools, checking its own work, then writing a final answer. The user sees one response. The vendor pays for every step in between.

That gap between what the user sees and what the vendor pays is now called the 100x problem. A query that costs a few cents on a chatbot can cost dollars on an agent system, sometimes much more.

Here is a concrete example. A user types: "What did our top customer ask about last week?" Simple question. But a typical agent touches seven separate billed operations to answer it: reading the user's message, loading its instructions and tool list, pulling in relevant records, choosing which tool to use, running that tool, summarising the results, and deciding whether to ask a follow-up. Together, those steps bill roughly 35,000 tokens. Tokens are the small chunks of text, usually a few characters each, that AI models count and charge for.

At current prices on a frontier model, that one query costs $0.10 to $0.40. Across one million queries a month, a typical target for any business software feature, the bill lands in the six figures.

Why does this matter to ordinary software customers?

It matters because the standard pricing model for business software breaks under this pressure. Most AI software today sells on a per-seat, per-month basis, meaning a company pays a flat fee for each employee who uses the tool. That works fine if costs per user stay predictable. Token amplification destroys the prediction.

A power user running 50 to 100 agent tasks a day on a $40-per-month plan can cost the vendor more in AI processing than that plan collects in revenue. The vendor loses money on its best customers. That is the paradox that, as VentureBeat reported this week, is now showing up in real company accounts.

Salesforce's Agentforce product is the most public example. A gap has opened between what the marketing shows and what is actually shipping to customers. That kind of gap appears when a feature is technically possible but too expensive to deliver at the price the subscription implies.

Nvidia VP Bryan Catanzaro put it plainly: "For my team, the cost of compute is far beyond the costs of the employees."

The fixes exist. Vendors can route cheap queries to smaller, cheaper models and save roughly 60% on inference bills. They can cache, meaning reuse, repeated instructions so the model does not re-read them from scratch on every call. They can trim the information passed to the model on each step. None of these are exotic. All of them require treating AI cost as a core engineering priority, not an afterthought.

For users of business software, the practical signal is this: if an AI feature your company pays for seems less capable than the demo promised, pricing pressure is a plausible explanation. Ask your vendor how they meter agent usage and whether heavy use triggers extra charges. The honest ones will have an answer ready.