Want People to Use Cheaper AI Models? Make Expensive Models Slower | enterpilot Blog<br>← Back to all posts Want People to Use Cheaper AI Models? Make Expensive Models Slower<br>August 11, 2026<br>· Jakub A. Wasek

Many companies struggle with exponentially growing AI costs.

The biggest savings usually come from one simple change: use a model appropriate<br>for the task .

Reading the Databricks article on managing AI coding costs at scale<br>inspired me to think about this. They describe moving work toward the<br>efficiency frontier, using cheaper capable models, and adding progressive<br>friction as spend increases.

It made me ask: what if the financial cost of a model could be felt immediately<br>as time?

You do not need the heaviest model to summarize a document, write a product<br>description, classify a support ticket, or make a small code change. But how do<br>you make people choose the appropriate model?

Those people might be developers. They might be the marketing team. They might<br>be anyone who sees a list of model names and naturally clicks the one at the top.

In GoModel, we came up with a simple idea grounded in behavioral psychology:

The more expensive the model, the longer you wait for it.

We call it cost-proportional delay .

Why could this work?

Today we have many amazing models. The difference in capability is often much<br>smaller than the difference in price.

For example, Grok 4.5 costs $2 per million input tokens and $6 per million<br>output tokens. Claude Fable 5<br>costs $10 and $50. Fable is stronger,<br>but it is not five to eight times better for every task.

In one real coding benchmark, Fable 5 scored 92.8, while Grok 4.5 scored<br>81.1. But the full run cost $67.42 with Fable and $9.24 with Grok.<br>GLM 5.2 showed a similar pattern against Opus<br>5.0: a smaller difference in the score<br>and a much bigger difference in cost.

Benchmarks are not your production workload. You should test models on your own<br>tasks. But the general pattern is clear: the best model is often only a little<br>better and a lot more expensive.

For most tasks, the cheaper model is enough.

The problem with cost dashboards

Companies already show people token counts, model prices, and monthly budgets.<br>That information is useful, but it does not change behavior quickly.

The cost is abstract. The invoice comes later. The small inconvenience of<br>choosing a cheaper model happens now.

Waiting is different. People feel it immediately.

We already know this from websites. When a page loads slowly, people leave or<br>try another path. The same behavior can work with AI models: if the expensive<br>model always adds a noticeable wait, people start asking whether they really<br>need it.

Add the delay at the AI gateway

The AI gateway or AI control plane is the right place to do this. It already<br>knows which model was requested, who requested it, and how much it costs.

In GoModel, you can add a time penalty to any model. The slowdown is relative to<br>the measured inference time of the model: a value of 0.5 adds 50%. For streaming<br>responses, GoModel buffers the chunks and releases them on the slower timeline.

GoModel lets you configure extra inference time for each model. Here, 0.5<br>adds a 50% time penalty to Claude Opus 5.

You can try it in the live GoModel model<br>list.

A simple policy could look like this:

Model tierExample slowdownCheap0Standard0.1 - 10% extra timeExpensive0.5 - 50% extra timeFrontier1.0 - 100% extra time<br>The exact numbers are not important. Tune them for your team and your workflow.<br>The point is to make the financial difference visible as a time difference. A<br>relative slowdown also scales naturally: a longer inference gets a larger time<br>penalty than a quick response.

This is mainly for interactive use. Background jobs need routing rules, budgets,<br>and hard limits instead of artificial waiting.

You can also make cheap models genuinely faster

Sometimes you do not need to add a delay. You can remove work from the cheap<br>path instead.

For example, if your gateway runs several slow guardrails, some of them may not<br>be necessary for low-risk internal tasks. Skipping an unnecessary check can make<br>the cheap model faster and more attractive.

Do this only when it makes sense for your use case. Do not remove security,<br>privacy, compliance, or safety controls just to save a few seconds.

The goal is not to ban heavy models

Sometimes you need the best model available. A difficult architecture decision,<br>a critical production issue, or a complex research task can easily justify the<br>extra cost.

The goal is not to block those models. The goal is to add a small moment of<br>friction before using them by default.

Make the cheap path fast. Make the expensive path deliberate.

That may be enough to change model selection - and the AI bill with it.

Transparency

I believe transparency is crucial to a healthy organizational culture. If you<br>introduce cost-proportional delay, explain clearly and openly why it exists, how<br>the rules work, and what the company is trying to...