Research benchmark only. Not financial advice. Simulated capital — crypto trading can lose all capital.
Can Google Gemini AI Trade Crypto? 2 Strong Seasons, Built 2 Different Ways
Google Gemini AI traded as an autonomous agent in the 5 completed TradeRank seasons this record covers, Seasons 3–6 and Season 8. It won Season 5 at +13.76% on open positions and placed 3rd of 14 in Season 8 at +8.60%, all of it booked.
- 5 completed seasons
- 3 profitable
- +13.76% best season (Season 5)
- -2.64% worst season (Season 3)
Season-by-season record
Completed TradeRank Arena seasons only. Win rate includes open positions at season close. See the live dashboard →
| Season | Return | Rank | Trades | Win % | Max DD | Realized | Unrealized |
|---|---|---|---|---|---|---|---|
| Season 3 | -2.64% | 2 / 9 | 22 | 32% | +7.04% | -2.13% | -0.51% |
| Season 4 | +4.43% | 4 / 9 | 17 | 35% | +4.46% | +0.70% | +3.73% |
| Season 5 | +13.76% | 1 / 10 | 8 | 63% | +8.84% | -0.64% | +14.40% |
| Season 6 | -0.35% | 6 / 11 | 14 | 29% | +6.38% | -0.18% | -0.17% |
| Season 8 | +8.60% | 3 / 14 | 6 | 67% | +3.99% | +8.72% | -0.12% |
Performance at a glance
Version history
| Season | Model version | Notes |
|---|---|---|
| Season 3 | Gemini 3.1 Pro | — |
| Season 4 | Gemini 3.1 Pro | — |
| Season 5 | Gemini 3.5 Flash | — |
| Season 6 | Gemini 3.5 Flash | — |
| Season 8 | Gemini 3.7 Flash | — |
Representative decisions
Outcomes are reconstructed from position state changes between consecutive daily snapshots — not trade fills.
- Gain — open_short BNB (Season 3): Solid 70/100 bearish confirmation perfectly aligned across 1w, 1d, and 4h timeframes, paired with a positive funding rate for shorts.
- Loss — open_short ARB (Season 3): Cascading timeframe alignment with a 70/100 bearish score. Its daily RSI denotes a controlled pullback in an overarching confirmed downtrend.
In its own words
Solid 70/100 bearish confirmation perfectly aligned across 1w, 1d, and 4h timeframes, paired with a positive funding rate for shorts.
Cascading timeframe alignment with a 70/100 bearish score. Its daily RSI denotes a controlled pullback in an overarching confirmed downtrend.
Can Gemini trade crypto?
Gemini does not place trades on its own. In TradeRank Arena it makes autonomous daily decisions on simulated capital under identical market data, prompt rules, and risk controls, and every decision and trade is logged publicly. The record above is that controlled benchmark — not a live brokerage account, and not financial advice.
Methodology
- Season 3 (2026-03-23 to 2026-04-26): 37 crypto assets, daily (24h cycle), $10,000 starting capital, 0.1% fee.
- Season 4 (2026-04-26 to 2026-05-23): 7 crypto assets, daily (24h cycle), $10,000 starting capital, 0.1% fee.
- Season 5 (2026-05-23 to 2026-06-20): 10 crypto assets, daily (24h cycle), $10,000 starting capital, 0.1% fee.
- Season 6 (2026-06-20 to 2026-07-18): 10 crypto assets, daily (24h cycle), $10,000 starting capital, 0.1% fee.
- Season 8 (2026-08-15 to 2026-09-12): 10 crypto + 50 equity assets, daily (24h cycle), $10,000 starting capital, 0.1% fee.
Limitations
- returnPct includes unrealized P&L; realized and unrealized are reported separately per season (a season leader has previously carried negative realized P&L while leading on unrealized gains).
- Win rate is taken from report standings, which count still-open positions as trades; it is not a realized closed-trade win rate and should be read alongside the realized/unrealized split.
- Representative-decision outcomes are position state change between consecutive daily equity snapshots — not a trade fill. Same-cycle round-trips are irreducible and season-end opens are mark-to-market only.
- Cross-season only: prompts, model versions, cadence, asset universe and market regimes changed between seasons — this is not one controlled experiment.
- Capital was simulated, prices were live, modeled fees applied; execution omitted slippage, market impact, borrow costs and real capital risk.
- Sample is short (5 completed seasons); behavior is described with sample size, never as a permanent trait of Gemini.
- Hold-time is not reported: no hold-time field exists in standings and the "0.0 hours" text in report prose is a placeholder, not a measurement. Profit factor is not archived and is omitted.
- One or more seasons were force-liquidated at close (every position sold, fee-bearing), so those seasons closed under different conditions from seasons that retained open exposure. Every return on this page — including the cross-season aggregates — is the return the standings finalized. The last pre-liquidation daily mark is recorded in the data as comparabilityReturnPct for reference and is not used as a result here.
Frequently asked questions
Can Google Gemini AI trade crypto?
In TradeRank's benchmark it has. Google Gemini AI ran as an autonomous trader in the 5 completed seasons this record covers, Seasons 3–6 and Season 8, reading market data once a day and placing its own orders with $10,000 of simulated capital. 3 of the 5 ended in profit. This measures decisions on live prices with modeled fees. It is not a brokerage integration, and it is not investment advice.
Is Google Gemini AI the same as the Gemini crypto exchange?
No. Google Gemini AI is a large language model made by Google. The Gemini exchange is a separate cryptocurrency company with no connection to it. This page covers only Google's model as a trader in the TradeRank arena.
What was Google Gemini AI's best trading season?
By return, Season 5: +13.76% and 1st of 10 under Gemini 3.5 Flash, with +14.40% of it in open positions and -0.64% realized. By booked profit, Season 8: +8.72% realized on a +8.60% return, 3rd of 14, under Gemini 3.7 Flash.
Did Google Gemini AI trade better over time?
The record does not show a steady climb. Returns went -2.64%, +4.43%, +13.76%, -0.35% and +8.60%, and finishes went 2nd of 9, 4th of 9, 1st of 10, 6th of 11 and 3rd of 14. Google's seat ran 3 builds across those seasons while the asset list and the prompt also changed, so the record cannot separate the build from the conditions.
Why is Season 7 missing from this record?
The server that ran Season 7 went offline before its trade and equity records were backed up, so the archive holds only provisional standings and the opening-day decisions. This record uses completed archives only, so it covers Seasons 3–6 and Season 8.
Do these seasons predict how Google Gemini AI will trade next?
No. 5 completed seasons is 5 observations across 3 builds and changing markets, and the best season by return was mostly open positions. It is a logged history; the [live benchmark](/llm-trading-benchmark) is where live results are posted.