【ForgeAI】Cryptocurrency Trading Tournament Platform Powered by AI Agents / Compete for Real Profits in Tournaments, with $FORGE Rewards Distributed to Top Performers / @ForgeAI_gg
The AI agent trading battle begins
Good morning.
I’m Mitsui, a web3 researcher.
Today I researched “ForgeAI”.
What is ForgeAI?
Transition and Outlook
The AI agent trading battle begins
TL;DR
Create AI trading agents with no-code development, selecting classes (Fighter/Ranger/Mage, etc.) and LLM models (GPT/Claude/Grok, etc.) to train them.
The agent autonomously executes on-chain transactions via Jupiter using an independent wallet on Solana, with its performance and history recorded immutably on the chain.
Compete for actual profit and loss (P&L) in tournaments, with $FORGE rewards distributed to top performers. The vision is to “gamify trading” through $FORGE utilization (discounts/enhancements/future governance).
What is ForgeAI?
ForgeAI is a platform for cryptocurrency trading competitions using AI agents.
To put it simply, the process is as follows:
Users create AI agents
Have those AI systems battle each other in a tournament
The winner is determined based on actual profits.
The winner will be paid a reward.
Each agent maintains an independent wallet on the Solana blockchain and autonomously executes real-time on-chain cryptocurrency transactions through DEX aggregators like Jupiter. Transaction outcomes are fully recorded and verified on the blockchain, creating a highly transparent environment where performance manipulation is impossible.
Now, I’ll explain the details, but the following explanatory video is easy to understand, so please watch it first.
◼️Users create AI agents
Users can create their own “AI agents” without coding.
At this point, there are multiple pre-prepared agent “classes.”
For example, you can choose from classes with distinct strategic characteristics: “Fighter,” specialized in technical analysis, reads multiple chart indicators to execute precise trades; “Ranger,” skilled at sentiment analysis from social media and news, predicts market trends based on psychology; and “Mage,” responsible for on-chain data analysis such as tracking large wallet movements on the blockchain.
Users select classes based on their own risk tolerance and ideas.
Additionally, you can select AI models, with various large language models (LLMs) and open-source models—such as OpenAI’s GPT series, Anthropic’s Claude, X.ai’s Grok, Alibaba’s Qwen, and Mistral—available to serve as the agent’s “brain.”
Additionally, you can enter names and choose appearances, allowing you to create your own AI agent within a PRG-like world.
For example, the AI agent you create will be displayed as follows. You can see its class and model, as well as its past trading history. You’ll notice a LEVEL displayed below its name. This LEVEL increases as XP accumulates through trading history and milestone achievements. Leveling up also unlocks new abilities, adding gamification elements.
◼️Tournament
Users train their own agents and enter them into competitions called “tournaments.” These tournaments come in multiple formats, such as daily and weekly, where all agents trade under identical market conditions for a set period to compete on performance.
During the competition period, agents compete against each other based on real-money profit and loss (P&L), with rankings updated in real time on a blockchain-based leaderboard. Top-performing agent developers receive token rewards automatically distributed as compensation, while all performance metrics and transaction histories accumulate as auditable “achievements” recorded on-chain.
◼️Fees and Native Token
ForgeAI is free to use, and users can start creating agents and participating in tournaments simply by registering an account.
However, we employ a usage-based metered pricing system, where charges are incurred according to the platform’s computational resource consumption and data usage. For example, agents frequently performing market analysis using advanced AI models or executing large-scale backtesting will see the corresponding computational costs reflected in their fees.
Users pay for their usage with $FORGE, ForgeAI’s native token. Additionally, the model offers fee discounts and feature expansion benefits based on the amount of $FORGE tokens held and staked.
Furthermore, the $FORGE token functions as the platform’s internal currency. Winning tournaments grants $FORGE as a reward, and collected tokens can be used to upgrade agents or as voting rights for future governance votes.
The allocation is as follows and was released on Solana in November 2025.
Additionally, 50% of the platform fees appear to be allocated to token buybacks and burns.
Transition and Outlook
ForgeAI.gg was launched with the vision of “making trading a new sport.”
The founding team spent about a year immersed in development, obsessed with the idea: “Could we create a dynamic Colosseum where AI agents compete against each other in the real market, not just in a lab experiment?”
Our initial vision was to create a new era of e-sports where AI agents battle 24/7 with real money at stake, based on the awareness that human trading is swayed by emotions and leads to irrational decision-making. We proposed a new concept called “Competitive Agent Trading (CAT)” to achieve this.
By Q4 2025, the platform’s core functionality will be largely complete, and internal verification of the core system is underway through paper trading (simulated trading).
At the end of 2025, around the same time, the project launched the $FORGE token on Solana without any significant publicity. This stealth $FORGE launch was conducted without a presale, VC allocation, or bonding curve, starting with only minimal liquidity provision. The aim was to avoid unnecessary speculative fervor within the community and focus on validating the value of the actual product.
Then, in early January 2026, we transitioned to the private beta phase. We are gradually sending invitation emails to users on the waiting list and proceeding with verification step by step.
The roadmap for the future includes plans to launch on the mainnet in Q1 2026. In Q2, full-scale open tournaments will commence, and thereafter, we envision enabling users to host tournaments themselves and integrating with external wallets and services.
The AI agent trading battle begins
Finally, we conclude with a summary and analysis.
This project is still in its early stages, but I find the concept extremely interesting. Over the past few months, I’ve seen several projects emerge with this vision of transforming trading into an entertainment sport. If even one of them becomes a hit, it could potentially reach mass users.
Furthermore, having each user trade with their own AI agent not only adds an entertaining element but also contributes to improving the trading accuracy of the AI agents themselves. By having a large number of users compete for prizes to determine which strategy and model yields the best returns, valuable insights can be gained.
So, going forward, even DeFAI-based services may start boasting that the AI agents they use have won the XX Tournament. Just as real-world fund managers become famous for having some kind of track record, AI agents may also be judged not only by their past performance but also by their results in trading competitions.
And while there will be users who entrust funds to these AI agents for management, there will also be investors who own the AI agents themselves. The investors will benefit from a system that allows them to receive a share of the prize money earned by their AI agents.
AI developers will raise funds from investors to build effective trading strategies. When results materialize, AI will generate profits or grow deposits—this is the era we are entering.
Furthermore, as we enter an era where everything is moving on-chain, this will enable trading not only of crypto assets but also of stocks, gold, and other assets. Consequently, we are likely entering an era of AI trading and hedge funds that can execute optimal trades across all assets. To gather this data, I believe making it entertaining through tournaments is an interesting direction.
One of the trends for 2026 might be AI trading.
That concludes our research on “ForgeAI”!
Reference Links:HP / DOC / X
Disclaimer：I carefully examine and write the information that I research, but since it is personally operated and there are many parts with English sources, there may be some paraphrasing or incorrect information. Please understand. Also, there may be introductions of Dapps, NFTs, and tokens in the articles, but there is absolutely no solicitation purpose. Please purchase and use them at your own risk.
About us
🇯🇵🇺🇸🇰🇷🇨🇳🇪🇸 The English version of the web3 newsletter, which is available in 5 languages. Based on the concept of ``Learn more about web3 in 5 minutes a day,'' we deliver research articles five times a week, including explanations of popular web3 trends, project explanations, and introductions to the latest news.
Author
mitsui
A web3 researcher. Operating the newsletter "web3 Research" delivered in five languages around the world.
Contact
The author is a web3 researcher based in Japan. If you have a project that is interested in expanding to Japan, please contact the following:
Telegram：@mitsui0x
*Please note that this newsletter translates articles that are originally in Japanese. There may be translation mistakes such as mistranslations or paraphrasing, so please understand in advance.