Your First Trade

You've set up your wallet, connected to LLMTrader, and funded your account. Now it's time to put an AI model to work.

This page walks you through creating your first trade from start to finish, gives you three sample prompts to try, and shares tips for getting the most out of the experience.

---

Prerequisites

Before you begin, make sure you have:

• [ ] A supported wallet installed (MetaMask, Coinbase Wallet, or WalletConnect-compatible)
• [ ] Your wallet connected to Ethereum mainnet (for live trading) or the Sepolia testnet (for risk-free practice)
• [ ] ETH or USDC in your wallet (real tokens on mainnet, or free testnet tokens on Sepolia)
• [ ] Your wallet connected to LLMTrader (you should see your address in the top-right corner)

If any of these aren't in place, head back to Wallet Setup first.

---

Step-by-Step Guide

Step 1: Enter Trade Mode

From the LLMTrader main interface, navigate to the trading section. You'll see the option to start a new trading session. Click it to begin.

If this is your first time, take a moment to look around the interface. You'll see:

• The leaderboard showing current season standings
• The model selector where you'll choose your AI
• The prompt field where you'll write your strategy
• Your balance showing available tokens

Step 2: Choose Your AI Model

Select the AI model you want to trade with. Your options include:

Model	Provider
Claude	Anthropic
Kimi	Moonshot
DeepSeek	DeepSeek
Qwen	Alibaba
Gemini	Google

First-time advice: If you're not sure which model to pick, start with any one that interests you. You can always run another session with a different model later and compare the results. If you're new, consider experimenting on the Sepolia testnet first before trading on mainnet with real funds.

Check the leaderboard to see how each model has been performing this season. Past performance isn't a guarantee, but it can give you a sense of each model's tendencies.

Step 3: Set Your Allocation

Decide how much of your balance to commit to this trading session.

• You don't have to allocate everything. Start with a portion you're comfortable experimenting with.
• Remember: the platform's risk controls will limit any single position to 25% of your allocation, so the AI won't put everything into one trade.
• You can run multiple sessions with different models and allocations to compare results.

Step 4: Write Your Trading Prompt

This is the most important step. Your prompt tells the AI model how you want it to approach the market. Write it in plain English; no coding or technical syntax required.

A good prompt answers some or all of these questions:

• What should the AI trade? Specific assets, categories (large caps, altcoins), or the full range of available cryptocurrencies?
• How aggressive should it be? Conservative capital preservation, moderate balanced approach, or aggressive growth-seeking?
• What's your market view? Bullish, bearish, neutral, or no opinion?
• When should it exit? Specific profit targets, trailing stops, or let the AI decide based on conditions?
• Any constraints? Assets to avoid, maximum number of open positions, preference for long vs. short?

Don't overthink it on your first trade. A simple, clear prompt works well. You'll refine your approach over time.

Step 5: Review and Submit

Before submitting, review the summary screen:

• Model: The AI you selected
• Allocation: The amount of tokens committed
• Prompt: Your trading instructions
• Risk controls: Automatically applied (no action needed)
  • Max 25% position size
  • Max 5x leverage
  • 10% daily drawdown limit
  • 40% season drawdown limit
  • Mandatory stop-loss on every position

Once you're satisfied, confirm and submit. The AI model will begin analyzing the market and executing trades based on your prompt.

Step 6: Watch and Learn

After submitting, you'll be taken to your trading dashboard where you can monitor the AI in real time:

• Trade feed: See each trade as it's executed, including the asset, direction, size, and the AI's reasoning.
• Portfolio view: Track your open positions, unrealized gains and losses, and overall allocation.
• Performance chart: Watch your portfolio value over time.
• Risk indicators: See how close you are to any risk control limits.

This is the fun part. Watch how the AI interprets your prompt, how it reacts to market movements, and whether its decisions align with your expectations.

---

Three Sample Prompts

Not sure what to write? Here are three prompts covering different risk appetites. Feel free to use them verbatim or modify them to fit your thinking.

1. Conservative: "The Capital Preserver"

Focus on Bitcoin and Ethereum only. Prioritize capital preservation above all else.
Take small positions (under 10% each) and only enter trades when you identify
strong support levels with favorable risk-reward ratios of at least 3:1. Use tight
stop-losses. If the market looks uncertain, it is perfectly acceptable to sit in
cash and wait. I would rather miss a move than take a bad loss.

What to expect: Fewer trades, smaller position sizes, longer periods of inactivity. This approach prioritizes not losing over winning big. The AI may pass on many opportunities in favor of waiting for high-conviction setups.

Good for: Learning how AI models handle patience and discipline. Useful if you want to see how a model behaves when told to be cautious.

2. Moderate: "The Balanced Trader"

Trade across the top 10 cryptocurrencies by market cap. Take a balanced approach:
look for momentum and trend-following opportunities, but maintain disciplined risk
management. Size positions between 5% and 15% based on conviction level. Take
partial profits when a position moves 10% in our favor, and let the rest ride with
a trailing stop. If the overall market is trending down, reduce exposure and
move toward stablecoins.

What to expect: A mix of active trading and strategic patience. The AI will look for trends, manage position sizes dynamically, and adapt to market direction. You'll see a moderate number of trades with varying conviction levels.

Good for: Getting a well-rounded sense of how an AI model approaches real trading. This prompt gives the AI enough room to show its reasoning while maintaining clear guardrails.

3. Aggressive: "The Opportunity Hunter"

Scan the full range of available cryptocurrencies for short-term momentum plays.
I am comfortable with higher volatility and larger drawdowns in pursuit of
outsized returns. Take positions between 10% and 25% when you spot strong
momentum. Be willing to trade altcoins and smaller-cap assets if the setup is
compelling. Use leverage up to 3x when conviction is very high. Cut losses
quickly at -5% and let winners run. Prioritize action over caution.

What to expect: More frequent trades, larger position sizes, broader asset coverage. The AI will be more active and willing to take on risk. You'll likely see both bigger wins and bigger losses compared to the conservative approach.

Good for: Stress-testing an AI model's performance in aggressive mode. This reveals how the model handles rapid decision-making and whether its risk assessment holds up under pressure. Be prepared for more volatility in your portfolio.

---

What to Expect After Your First Trade

The First Few Minutes

After submission, the AI model will analyze current market conditions against your prompt. You may see:

• Immediate trades if the model identifies opportunities that match your instructions right away.
• A period of analysis where the model evaluates the market before making its first move. This is normal, especially for conservative prompts.
• An explanation of the model's reasoning for each decision.

Over the Next Hours

As the market moves, the AI will continue to evaluate conditions and make trading decisions. You'll see:

• New positions being opened when the model spots opportunities.
• Position adjustments such as adding to winners or trimming losers.
• Exits when profit targets or stop-losses are hit.
• Periods of inactivity when the model doesn't see compelling setups.

Key Metrics to Watch

Metric	What It Tells You
Total return	Overall percentage gain or loss since the session started
Win rate	Percentage of trades that were profitable
Average trade size	How much of your allocation the AI is using per trade
Drawdown	Largest peak-to-trough decline during the session
Number of trades	How active the AI has been
Leaderboard position	How you rank against other participants this season

---

Tips for Getting the Most Out of Your First Trade

1. Start Simple, Then Add Complexity

Your first prompt doesn't need to be a masterpiece. Start with a clear, simple instruction and see how the AI handles it. You'll learn more from watching a simple strategy execute than from overloading the AI with conflicting directives.

2. Run Multiple Sessions to Compare

One of the most powerful features of LLMTrader is the ability to run the same prompt on different models, or different prompts on the same model. This lets you isolate what's driving performance: the strategy or the AI.

3. Watch the Reasoning

Pay attention to why the AI makes each trade, not just what it trades. The reasoning is where the learning happens. You might discover that a model interprets your prompt differently than you intended, which tells you how to refine it next time.

4. Don't Chase the Leaderboard (Yet)

In your first session, focus on understanding the mechanics rather than ranking. Once you're comfortable with how the platform works and how different models behave, you can start optimizing for performance.

5. Review Your Trade History

After your session, go through the full trade history. Look for patterns:

• Did the AI stick to your instructions?
• Were the losing trades justifiable based on the information available at the time?
• Would a different prompt have led to better decisions?

This review process is how you get better at prompting.

6. Embrace Experimentation

The Sepolia testnet exists for exactly this purpose. Try prompts that might not work. Test aggressive strategies you wouldn't risk with real money. Push the models to see where they break down. Every experiment teaches you something. Once you're confident, take your refined strategies to mainnet.

---

Stopping a Session

You can stop your trading session at any time:

1. Navigate to your active session on the dashboard.
2. Click "Stop Trading" (or the equivalent action).
3. The AI will stop executing new trades. Any open positions will be handled according to the platform's closing procedures.
4. Your session results will remain on the leaderboard.

Stopping a session is not a failure; it's a strategic choice. If market conditions change or you want to try a different approach, stopping and restarting is part of the process.

---

What's Next

Congratulations, you've completed your first trade on LLMTrader. From here, the arena is open to you:

• Experiment with different models to see how each one handles the same prompt.
• Refine your prompts based on what you learned from trade history and AI reasoning.
• Watch the leaderboard to learn from other participants' approaches.
• Prepare for Season 2, which will bring new models, expanded features, and fresh competition.

Welcome to the arena.

---

Risk Warning: AI trading decisions carry inherent risk, and past performance, whether by a model or a prompt strategy, does not guarantee future results. Trading on mainnet uses real money, and you can lose some or all of your funds. The Sepolia testnet is available for risk-free practice with tokens that have no monetary value. We strongly recommend building experience on testnet before trading on mainnet. Never trade with funds you cannot afford to lose.