Data Insight
Jul. 2, 2026

GPT-4 led in ECI far longer than any other model

Jaeho Lee's avatar
By Jaeho Lee

OpenAI’s GPT-4 topped the Epoch Capabilities Index (ECI), our aggregate measure of language model capability, for roughly a year after its release in March 2023. No model since has led for as long. The second-longest lead, by OpenAI’s o1, lasted a little over three months, less than a third of GPT-4’s.

The chart above shows the ECI frontier over time, where each step marks a new model taking the top spot. Claude 3 Opus overtook GPT-4 in February 2024, the first of 17 leading models through June 2026, with each holding the lead for a median of about seven weeks. We begin the comparison at GPT-4’s release date, March 14, 2023.

Epoch's work is free to use, distribute, and reproduce provided the source and authors are credited under the Creative Commons BY license.

Learn more about this graph

We track which model has the highest ECI at each point in time, and how long it stays at the top of the leaderboard. The ECI can fluctuate slightly over time, though rarely enough to significantly shuffle model rankings. That said, this is a retrospective view and does not account for changes to ECI estimates since models’ release dates.

Data

Analysis

Assumptions and limitations

Download this data

Benchmarked models with Epoch Capabilities Index (ECI) scores
CSV, Updated Jul. 2, 2026

Explore this data

AI Capabilities
AI Capabilities

Benchmark results featuring the performance of leading AI models on challenging tasks.

Chinese AI models have lagged the US frontier by 7 months on average since 2023
Data Insight
Jan. 2, 2026
Chinese AI models have lagged the US frontier by 7 months on average since 2023
Open models lag state-of-the-art closed models by 4 months
Data Insight
May 29, 2026
Open models lag state-of-the-art closed models by 4 months

Related topics

Capabilities