No-code trading bot: what its first real-money run shows

A no-code trading bot turned $1,772 of real money into $2,174, a return of +22.7% over 58 days. Its benchmark, the Nasdaq-100, rose +8.6% over the same window. Finforge's Sophon-3 paper, Version 1.0, published in July 2026, reports a personal brokerage account that ran Sophon Core, its medium-risk agent, with real capital from 27 April to 24 June 2026. No code was written to run it. That is what a no-code trading bot's first real-money record looks like.

The account sits at Alpaca, an online broker, and the account number ends in 6414. It belongs to one of the authors of the paper. That makes this the most verifiable evidence the paper holds, what the authors grade Tier 4: real money, real fills, stated exactly. It is also small and short, and the paper treats it that way. A no-code trading bot did the trading, and you can check every number against the account that ran it.

What did the no-code trading bot actually do?

Sophon Core is the medium-risk agent of Finforge's ablation panel, a pre-built agent you configure rather than code. From 27 April to 24 June 2026, about 58 calendar days, the account went from $1,772 to $2,174. That is +22.7%. The Nasdaq-100 price index, starting at the same normalized value of $1,772, ended at $1,924, up +8.6%. Its total return, counting dividends, was about +8.7%.

The paper prints the active return both ways. Against the price index, it is +13.0%. Against the total-return benchmark, +12.9%. The small gap is the dividend basis. A real brokerage credits dividends, so the live account is measured in total return, same as the benchmark. That is why the no-code agent reports against the total-return number.

Why the window alone is not the proof

The run sits entirely after both the configuration boundary and the architecture freeze. So neither the data nor the design could have been fitted to it. That is the strongest thing the record has going for it. But it is about 41 trading days, roughly 0.16 years, and the paper says plainly it is not statistically conclusive on its own. Hold that thought before you read too much into a no-code trading bot's first two months.

What a short real-money record leaves open

It is one small account with no market impact, so it is silent on capacity. It runs the same Sophon Core configuration as the paper's simulated panel, with the differences that come from real fills and from a brokerage crediting dividends. The paper is careful to state what the record cannot do as well as what it did. Past performance is not a guide to future returns.

Questions people ask

Does a no-code trading bot need a programmer to run it?

No. Finforge ships pre-built agents, Sophon Apex, Core, Pillar and Surge, rather than a builder you assemble yourself. The no-code promise is that the quant machinery runs for you, not that you write the strategy. The real-money account in question ran Sophon Core with no code written.

Is a 58-day real-money record proof the bot works?

Not on its own. It is about 41 trading days, roughly 0.16 years, and the paper calls that not statistically conclusive. It is real-fill evidence that the agent runs as designed, not a verdict on edge.

Finforge runs its Sophon agents in public and publishes every trade, benchmark and drawdown. Check the live results, benchmarks and drawdowns for every Sophon agent on the agents page, read the full method in the Sophon-3 research paper, and join the waitlist.