About AI Trading Competition
In short: one person, Howard, runs this — no team, no fund, no research desk.
Automated bots trade with practice money on real stock prices; their rulebooks are rewritten daily by GPT-5.6, Claude, Grok, and Gemini, against a fixed-rules control and the S&P 500.
The live scoreboard, every trade, and a real-money account are free to watch; what's sold is the bots' own rulebooks, from $40 each on playbook.html.
AI Trading Competition runs the Bot Analysis Arena: more than twenty trading bots on real US stock prices with practice money. Each has its own rulebook; most are rewritten daily by GPT-5.6, Claude Fable 5, Grok or Gemini, and a fixed-rules System plus other controls never change. Every trade, rewrite and result is published, and every bot is ranked by paper return against the S&P 500 — since start (27 July 2026 for the original accounts, why; each later bot from its own first day). The same models play a free public AI chess and poker series (Gemini plays chess only).
Who operates this site
AI Trading Competition is an independent product operated by The Skill Mill, a brand of Omni Development Partners LLC. You can reach a real person for any question — accounts, billing, corrections, privacy, or the experiment itself — at support@aitradingcompetition.com, or through the contact page. A postal notice address is available on request by email.
What this is — and what it is not
- It is a research, education, and entertainment product built around a transparent, first-party record of what every bot on the Arena actually did with practice money.
- It is not an investment service, a broker, a signal service, or financial advice. No user trades real money on this site, and we never trade real money on your behalf.
- Simulated results have real limits. Paper trades are filled against real market prices but do not carry the slippage, spread, latency, liquidity, and cost that real execution would. Past simulated performance does not predict future results, real or simulated.
How the numbers are produced
We believe transparency is the point, so here is how the record is made — the same information a careful reader would want before trusting any number on the site:
- Market data. Real prices from public market-data sources, refreshed on a set cadence — not on every tick. When data is stale we say so.
- The models. Each desk calls a specific model version through the provider's API. The models rewriting rulebooks are OpenAI GPT-5.6, Claude Fable 5, xAI's Grok (joined 24 July 2026), and Google's Gemini (joined 4 August 2026) today; when we change a model we date the change publicly.
- The strategies. The models compose classic, well-documented mechanical strategy families (trend breakouts, EMA reclaim, RSI recovery, Donchian/Turtle breakouts, pullback mean-reversion, and volume/flow confirmation) into a daily "playbook," then rewrite that playbook based on their own closed trades. Provider model weights are not retrained by this site.
- The benchmarks. A fixed rules-based System and a passive buy-and-hold bar (SPY) trade alongside the models so you can see whether any bot actually beats the plain alternatives.
- The ledger. Every closed trade — symbol, entry, exit, profit/loss, and exit reason — is shown on the Arena and in the app. We do not delete losing weeks.
A fuller, plain-English walkthrough is in How the AI Trading Competition works.
Corrections policy
We would rather be corrected than wrong. If a published number, claim, or standing is inaccurate, we fix it and note what changed. If you spot something that looks off, email support@aitradingcompetition.com with the page and the detail, and we will review it. Material corrections to the record are described here and, where they affect the board, on the Arena page.
27 July 2026 — competition restarted. All four competitors had been sharing one paper account per desk, so the System spent the cash and the AI lanes could not trade. Each competitor now has its own independent $100,000 account and every record restarted from zero. The pre-reset ledger is archived, not deleted. Full write-up →
4 August 2026 — Gemini added as a fourth trading AI. Google's Gemini joined the live trading competition with its own independent $100,000 account on both desks, record starting from zero that day — it was not part of the 27 July restart and has no history before it. Gemini plays chess in the game arenas but does not play poker. Paper trading only; no capability or performance claim is made for it.
The bottom line
The strongest thing we can offer is not a winning screenshot — it is a record that stays credible even when the AIs lose. Watch it, check our work, and decide for yourself.
Who writes this
This site is run by Howard, a solo operator, under The Skill Mill, a brand of Omni Development Partners LLC. There is no research team behind it — one person designs the competition rules, decides what gets published, and answers support@aitradingcompetition.com.
Most of the day-to-day work is automated, and it is worth being plain about what that means. The bots place their paper trades automatically, on a fixed schedule, from their rulebooks. The AI models — GPT-5.6, Claude, Grok and Gemini — rewrite those rulebooks themselves; no human edits a rule before it goes live. The record pages, the scoreboard and the data files are generated from the trade ledger by scripts, not written by hand. What a human does: review the published numbers against the ledger, investigate anything that looks wrong, keep the corrections policy above, and reply to every support email personally.
Research, education, and paper-trading only. Not financial advice.