Models & Research

OpenAI's agents now finish almost half of its 2 to 4 hour research tasks alone, and the median researcher uses $2,400 of AI a day

OpenAI's new research acceleration report says agent success on 2 to 4 hour research tasks rose from 23% to 46% since March, while median daily agent use per researcher hit $2,400 at list prices.

Researchers seen from behind at a whiteboard covered in abstract diagrams, no readable text
AI-generated

OpenAI has published its own numbers on how much AI is now doing AI research. “We find researchers’ AI use continues to rise exponentially, and agents are completing longer tasks, but human involvement remains central in research,” the company posted on X on Wednesday evening.

The headline figures

In the report, titled “Research acceleration: The jagged frontier”, OpenAI says estimated agent success on tasks that would take a human two to four hours rose from 23% in March to 46% in October. For four to eight hour tasks it rose from 13% to 35%.

The gains are uneven, which is where the “jagged” title comes from. OpenAI says autonomous agents succeeded 79% of the time at monitoring training runs, 35% at launching training runs and 25% at writing code for inference and serving stacks.

Background

Earlier stories on this thread.

What happened next

Later stories on this thread.

The spending numbers

Measured at API list prices, OpenAI says the median researcher’s daily agent usage reached $2,400 in October, up 62% from September. At the 90th percentile it reached $23,000 a day, up 105%. OpenAI notes its researchers do not currently have agentic coding quotas.

It also asked researchers what they would give up to keep frontier coding agents for three months. The median answers were 30% of team compute or, in a separate question, 40% of team headcount.

More people on safety

OpenAI says documented workweeks on safety and alignment rose 77% from January to September, the largest absolute increase of any of its major research areas, through new hires and researchers moving across. It says growth picked up after its response to what it calls the Hugging Face incident in July 2026.

Why OpenAI is publishing this

The company says AI labs should be required to track and publicly report indicators of recursive self-improvement, where AI speeds up the development of future AI, and stresses that the report “should not be read as an intention to pursue full RSI”. More on the company is on our OpenAI topic page.

Our read

Analysis: these are internal measurements from the lab that benefits from the story, and success rates are estimates. The useful part is the breakdown: agents look strong at watching runs and weaker at writing serving code, so humans still own the hardest steps.

This report is not legal, financial, or health advice. See the disclaimer.

Primary sources

  1. OpenAI on X: new data on how AI is changing research at OpenAI OpenAI on X
  2. Research acceleration: The jagged frontier OpenAI

Tags