GPT-4 held the Epoch Capabilities Index top for 52 weeks. Since Claude 3 Opus unseated it in February 2024, the lead has changed 17 times with a median stay of seven weeks.
Key facts
- GPT-4 held ECI top for ~1 year.
- Median top-model stay: 7 weeks.
- 17 leadership changes since Feb 2024.
- o1 held second-longest lead at 3 months.
- Competition fiercer but capability gains shrinking.
OpenAI's GPT-4 held the top spot on the Epoch Capabilities Index (ECI) for about a year, according to The Decoder. The ECI is a composite measure of language model performance. Epoch AI researcher Jaeho Lee noted that GPT-4's reign was far longer than any model since.
OpenAI's o1 held the second-longest lead at just over three months — less than a third of GPT-4's duration. Since Claude 3 Opus dethroned GPT-4 in February 2024, the lead has changed hands 17 times. The median stay at the top per model was about seven weeks.
The chart can be read two ways. It shows how long rival labs needed to match GPT-4, a genuine outlier at launch. But it also shows that competition has gotten much fiercer since. No model can hold a comparable lead anymore, and the capability jumps between transitions are faster but smaller compared to GPT-4 and the era that began with reasoning models like o1-preview in fall 2024.
The Shrinking Lead Window
The compressed leadership cycles reflect a market where frontier labs — OpenAI, Anthropic, Google, and others — release models on a cadence measured in weeks, not quarters. GPT-4 launched in March 2023 and wasn't surpassed until February 2024. Today, a model that tops the ECI in January may be obsolete by March. The Decoder reports that competition is fiercer, but the capability gains between transitions are shrinking.
What This Means for Buyers
For enterprise teams evaluating which model to build on, the churn creates a dilemma. Locking into a specific model architecture risks rapid obsolescence. The ECI data suggests that no single model will dominate long enough to justify a proprietary integration without a multi-model abstraction layer. Labs like Anthropic, with Claude Code and Claude Opus 4.6, and OpenAI, with GPT-4o and GPT-5.6, are racing to ship features — but the window to capture developer mindshare is narrowing.
Key Takeaways
- GPT-4 dominated the ECI for a year.
- Today's top models last 7 weeks median, with 17 leadership changes since Feb 2024.
What to watch
Watch for the next ECI update to see if any model can break the 7-week median. The release of GPT-5.6 or Claude Opus 4.6's successor could reset the clock — or confirm that the era of long-lived leaders is over.
Source: the-decoder.com
