Command Palette
Search for a command to run...

Open-Weight AI Models Lag Proprietary Models by Four Months, Epoch AI Finds

aiai-modelingai-research-evalsai-open-models 3 posts · 2 accounts

Epoch AI said its latest comparison of open-weight and proprietary AI models showed open-weight systems have lagged the state of the art by about four months since the start of 2026. The research group said the average gap between open and proprietary capabilities since January was 8 ECI points.

Epoch AI said that difference is similar to the gap between GPT-5 and GPT-5.5, providing a benchmark for how far open-weight models remain behind proprietary leaders.

From the sources (3 posts)

@epochairesearch

We took another look at the capability gap between open-weight and proprietary models. Since the start of the year, open-weight models have lagged the state of the art by four months.

@epochairesearch

The average gap between open vs. proprietary capabilities since Jan 2026 has been 8 ECI points — similar to the difference between GPT-5 and GPT-5.5. More on this Data Insight:

@kimmonismus

According to research by EpochAI, open-weight models lag behind frontier closed-source models by four months. Four months. That's very little. And impressive at the same time.

Preview built on a synthetic news corpus (16 weeks, Apr–Jul 2026). Impact calls are model reads, not price data.

About Archive