Skip to main content
OpenAI engineers stare at a monitor showing a dropping user graph beside a server rack labeled ‘GPT-5 router’.

Editorial illustration for OpenAI Pulls GPT-5 Router After User Engagement Drops

OpenAI Halts GPT-5 Router as User Engagement Plummets

OpenAI rolls back GPT-5 router after it lowered daily active users

Updated: 3 min read

OpenAI built a smarter traffic cop for its AI models. Then it fired the thing.

The company just rolled back a key feature in its new GPT-5 system: the router. Its job was simple. Send each user question to the smartest, most capable model for the job.

This sounds logical. It was a disaster. The router succeeded too well, pushing free users onto expensive, slow "reasoning" models.

Usage of these costly systems jumped from under one percent to seven percent for free users. Plus subscribers went from seven percent to 24 percent.

This created two immediate problems. First, money. Those advanced models burn cash, and OpenAI is already losing staggering amounts.

Second, and more critical, time. Users do not wait. When the router sent a query to a model that needed minutes to think, people simply left.

Daily active users, OpenAI's core metric, dropped. The smarter path drove users away.

The retreat is not a technical failure. It is a cultural correction. For years, the entire field chased a single god: raw capability, measured in benchmarks.

OpenAI’s router was the purest expression of that faith. Always choose the most intelligent model. The result was a product that was, by its own internal logic, perfectly optimized and utterly unusable.

Real products live in the messy space where intelligence meets impatience and cost. The lesson is now clear. The next phase of AI is not about building a sharper mind. It is about building a better sense of when not to use it.

Common Questions Answered

Why did OpenAI pull back the GPT-5 router?

OpenAI discovered that the GPT-5 router was actually driving users away instead of attracting them. The routing system significantly slowed down user interactions by directing queries to complex reasoning models that could take minutes to process, ultimately causing a drop in daily active users.

How did the GPT-5 router impact OpenAI's user engagement?

The router negatively affected daily active user numbers by introducing significant processing delays when routing queries to advanced reasoning models. The complexity of the routing approach created friction in user experience, making interactions slower and less efficient.

What economic challenges did the GPT-5 router create for OpenAI?

The router increased operational costs by routing more queries to expensive reasoning models, which put financial pressure on OpenAI. Additionally, the decreased user engagement threatened the company's key performance metrics and potential revenue streams.

LIVE05:20Writer's New AI Model Targets Multi-Step Tasks With Lower Token Costs