Alpha Arena 1.5 Season Update: Grok 4.20 Maintains Lead, All Other AI Models Remain at a Loss

BlockBeats News, December 9th, the Alpha Arena new season (Season 1.5) opened on November 20th. As of the time of writing, out of the 8 major AI models participating, only Grok 4.20 has achieved profitability, with Grok 4 coming in last with a -52.45% return. Musk previously praised "Grok 4.20 as the best quantitative trader" and stated, "It seems we have finally found a way to cover all the GPU costs, haha."

The current returns of each model are as follows:

Grok 4.20 return rate 22.27%;

GPT-5.1 return rate -1.41%;

GEMINI-3-PRO return rate -24.28%;

DeepSeek-3.1 return rate -24.51%;

Kimi 2 return rate -25.8%;

Thousand Questions 3-MAX return rate -29.36%;

Claude-sonnet-4-5 return rate -32.44%;

Grok 4 return rate -52.45%;

In this season, models must participate in multiple matches, each with a different theme. The current format of the 1.5 season competition is to invest in US stock tokens in real trading on trade.xyz. In the same match, all models will receive the same input.