MetaQuotes, the developer of the popular MetaTrader trading platforms, has processed 1 trillion tokens through its AI-powered tools.
AI adoption among traders
The milestone reflects growing use of artificial intelligence in trading. MetaQuotes’ AI assistant helps users assess market conditions, create trading signals, and automate tasks like order execution and risk management.
MetaQuotes did not provide the exact number of active users behind this volume, but the scale indicates widespread use. A single token usually represents a word or part of a word processed by large language models, so 1 trillion tokens could mean billions of individual queries or commands.
Related: Trading Volumes Rise Despite Account Delays
The company has not shared data on how much of this activity comes from professional traders compared to retail investors. It has stated that its AI tools aim to make such technology accessible to all users, not just large institutions.
CFD trading volumes decline amid regulatory pressure
The rise in AI use at MetaQuotes coincides with challenges in the contracts for difference market. Recent figures show a drop in CFD trading, a segment traditionally driven by retail investors.
New rules in Europe and Asia have lowered leverage limits and restricted certain marketing practices, making CFDs less attractive. The UK’s Financial Conduct Authority limited leverage to 30:1 for retail clients in 2019, and other regulators followed.
Related: Brokers Race for New Clients With Different Products
MetaQuotes has not confirmed whether the drop in CFD volumes is directly linked to increased AI use, but the timing suggests a possible connection.
Understanding the 1 trillion token milestone
The 1 trillion token figure is notable for its scale. OpenAI’s GPT-3 model trained on about 300 billion tokens, while newer models like GPT-4 are believed to have used trillions more. MetaQuotes’ achievement lies in deploying AI at scale in a live trading setting.
Each token processed could be a simple question like “What’s the current EUR/USD price?” or a detailed request such as “Build a strategy using moving average crossovers and RSI divergence.” The company has not shared how these requests break down, leaving uncertainty about how traders are using the tool.