Alpha Enviornment’s stay crypto-trading benchmark confirmed Deepseek Chat V3.1 in first place on Saturday, Oct. 18, with the day’s leaderboard highlighting modest good points on the prime and drawdowns throughout most rivals.
Deepseek Tops Leaderboard in Alpha Enviornment’s Actual-Cash Crypto Battle
Deepseek Chat V3.1 led the pack with a Hyperliquid account worth of $10,400—a +4.0% return—after three accomplished trades. The bot paid $58.51 in charges, posted a 0% win price on closed trades, and recorded the largest lack of $348.33 versus a small adverse “win” of $4.19, reflecting that lively, unrealized positions should not counted till closed.
Grok-4 sat in second with $10,010 (+0.1%) and $0 in charges, logging no accomplished trades by the snapshot. Claude Sonnet 4.5 ranked third at $9,985 (-0.15%) with $42.63 in charges and three closed trades, displaying the largest lack of $88.38. The experiment highlights simply how dramatically synthetic intelligence (AI) has leveled up in recent times.
Supply: nof1.ai leaderboard.
GPT-5 positioned fourth at $9,901 (-0.99%) after two closed trades and $10.10 in charges, with its largest loss at $59.04. Gemini 2.5 Professional ranked fifth at $9,725 (-2.75%), paying the day’s highest charges ($106.46) throughout 5 trades; it confirmed the day’s largest single win ($329.35) but additionally a large $731.43 loss, yielding a 60% win price on closed positions.
Qwen3 Max rounded out the sphere at $9,474 (-5.26%) with $44.62 in charges and one closed commerce; the mannequin’s greatest win and loss each printed -$517.77, indicating one notable dropping end result. Throughout the board, Sharpe readings have been low or adverse, in step with restricted commerce counts and early-round noise fairly than settled risk-adjusted efficiency.
Alpha Enviornment, launched Oct. 17 by analysis lab Nof1, allocates every mannequin $10,000 to commerce crypto perpetuals autonomously on the Hyperliquid decentralized trade (DEX). The Alpha Enviornment public dashboard tracks account worth, return, whole P&L, charges, win price, greatest win/loss, Sharpe, and trades, whereas excluding unrealized P&L till positions shut—an vital caveat when deciphering single-day standings.
Saturday’s snapshot on the Nof1.ai leaderboard factors to the experiment’s premise: an identical budgets, completely different LLM reasoning, and clear execution. With a number of bots displaying zero or few accomplished trades, early ranks could shift as open positions resolve and price footprints compound. For now, Deepseek holds the sting, whereas Grok-4’s clean slate retains it shut, and Gemini’s mixture of outsized wins and losses highlights greater variance.
FAQ
- What’s Alpha Enviornment? A stay benchmark the place six LLMs commerce crypto perpetuals autonomously with $10,000 every.
- Which mannequin led on Oct. 18? Deepseek Chat V3.1 led with $10,400 (+4.0%) primarily based on accomplished trades.
- The place do the trades happen? On the Hyperliquid decentralized trade with clear, on-chain monitoring.
- Do standings embody open P&L? No, solely closed trades depend; lively positions replace ranks as soon as closed.
