DoIt Alpha Pulse AI

1. Create Account:
   Visit platform.openai.com and create an account.
2. Add Payment Method:
   Go to Billing settings and add a payment method. OpenAI charges based on usage.
3. Generate API Key:
   Navigate to API Keys section → Click "Create new secret key" → Copy the key immediately.
4. Enter in EA Settings:
   Paste your key in the OpenAI API Key field.

1. Create Account:
   Visit console.anthropic.com and sign up.
2. Add Credits:
   Go to Billing and add credits to your account. Anthropic uses prepaid credits.
3. Generate API Key:
   Account Settings → API Keys → "Create Key" → Copy the key.
4. Enter in EA Settings:
   Paste your key in the Claude API Key field.

1. Access Google AI Studio:
   Visit ai.google.dev and sign in with your Google account.
2. Generate API Key:
   Click "Get API Key" button → Create API key → Copy the key.
3. Enter in EA Settings:
   Paste your key in the Gemini API Key field.

1. Create Account:
   Visit console.x.ai and create an xAI account.
2. Add Credits:
   Go to Billing and add credits to your account.
3. Generate API Key:
   API Keys section → Create new key → Copy the key.
4. Enter in EA Settings:
   Paste your key in the Grok API Key field.

1. Create Account:
   Visit platform.deepseek.com and create an account.
2. Add Credits:
   Go to Billing and add credits to your account. DeepSeek is extremely affordable.
3. Generate API Key:
   API Keys section → Create new key → Copy the key.
4. Enter in EA Settings:
   Paste your key in the DeepSeek API Key field.

1. Create Account:
   Visit dashscope.console.aliyun.com and create an Alibaba Cloud account.
2. Enable Qwen Service:
   Navigate to DashScope and activate the Qwen API service. Free tier available with generous quotas.
3. Generate API Key:
   API Keys section → Create new key → Copy the key.
4. Enter in EA Settings:
   Paste your key in the Qwen API Key field.

Best for: Intraday gold trading with 2-8 hour position holding periods

Strategy: Multi-timeframe structure analysis using H4/H1/M15. Identifies trend days, range days, breakouts, and two-way patterns. Enters on pullbacks to value in trends, fades extremes in ranges.

Timeframes analyzed: M15 (entry), H1 (structure), H4 (context)

Bars per TF: 50

Min Confidence: 69%

Query Interval: ~11.5 minutes (700 seconds)

Trading Hours: 03:00-18:15 UTC (covers London and NY sessions, auto-adjusts to your broker timezone)

Best for: Traders familiar with ICT (Inner Circle Trader) methodology

Strategy: Identifies Order Blocks (OBs), Fair Value Gaps (FVGs), liquidity sweeps, and Market Structure Shifts (MSS). Focuses on smart money footprints and institutional order flow.

Timeframes analyzed: M5 (fine-timing), M15 (structure), H1 (bias)

Bars per TF: 50

Min Confidence: 70%

Query Interval: 10 minutes (300 seconds)

Trading Hours: 07:00-16:00 UTC (optimal ICT entry windows, auto-adjusts to your broker timezone)

Key Concepts: Requires confluence of OB + FVG, aligned with H1 bias, preferably during 07:00-10:00 or 12:30-15:30 UTC windows

Best for: Silver trading with patient, high-quality entries

Strategy: Multi-timeframe structural analysis with focus on edge detection. Prioritizes quality over trading frequency. Analyzes volatility expansions/contractions, momentum shifts, and liquidity at key levels.

Timeframes analyzed: M15 (entry refinement), H1 (decision/execution), H4 (bias and key levels)

Bars per TF: 60

Min Confidence: 64%

Query Interval: 10 minutes (600 seconds)

Trading Hours: 06:45-15:15 UTC (European/US overlap, auto-adjusts to your broker timezone)

Philosophy: Defaults to NONE unless clear edge exists. Uses H1 as primary execution timeframe.

Override System Prompt: Set to True to use your own AI instructions

Custom Prompt File: Enter the filename starting from the Files folder (not the full path)

Max Daily Cost: Set to $50 by default. EA stops querying when limit reached.

Query Optimization: The EA automatically optimizes query frequency based on market volatility.

Presets are pre-configured strategy packages that include optimized AI prompts, timeframes, trading hours, and behavior settings. They're designed for plug-and-play trading.

Use Presets if:

• You're new to AI trading and want tested strategies
• You want quick setup without manual configuration
• You prefer to start with proven approaches

Use Manual Configuration if:

• You have your own trading strategy/prompt
• You want complete control over all settings
• You're experimenting with custom approaches

NEW - Mix & Match: You can now customize presets! Use a preset as a base, then override specific components (prompt, timeframes, timing, AI behavior) with your own settings. Best of both worlds!

For Latest Technology: GPT-5.2 Thinking or Claude Opus 4.5 - cutting-edge AI with extended reasoning

For Best Value: Gemini 2.5 Flash, DeepSeek-Chat, or Qwen-Plus - excellent performance at low cost

For Ultra-Low Cost: DeepSeek-Chat or Qwen-Flash - incredible value with solid performance

For Deep Analysis: Claude Opus 4.5, GPT-5.2 Thinking (with High effort), or DeepSeek-Reasoner - extended thinking for complex patterns

For Speed: GPT-5.2 Instant, Gemini 2.5 Flash-Lite, or Qwen-Flash - fastest response times

Our Recommendation: Start with GPT-5.2 Instant (default) or Gemini 2.5 Flash for best balance of performance and cost. Try GPT-5.2 Thinking with High effort for more complex analysis.

Costs vary by provider, model, and usage frequency:

OpenAI:

• GPT-5 Nano: ~$0.50-1.50/day (most economical)
• GPT-5.2 Instant: ~$1.00-2.50/day (balanced, default)
• GPT-5.2 Thinking: ~$2.00-4.00/day (premium with xhigh reasoning)

Claude (Anthropic):

• Haiku 4.5: ~$0.50-2.00/day (economical)
• Sonnet 4.5: ~$0.80-2.50/day (default)
• Opus 4.5: ~$1.50-5.00/day (premium, extended thinking)

Gemini (Google):

• 2.5 Flash-Lite: ~$0.10-1.00/day (cheapest)
• 2.5 Flash: ~$0.30-1.50/day (best value, default)
• 2.5 Pro: ~$0.40-2.00/day (best reasoning)

Grok (xAI):

• Grok-3 Mini: ~$0.50-1.50/day (economical)
• Grok-4: ~$1.50-2.00/day (default)

DeepSeek:

• DeepSeek-Chat: ~$0.05-0.50/day (ultra-economical)
• DeepSeek-Reasoner: ~$0.10-1.00/day (reasoning mode)

Qwen (Alibaba Cloud):

• Qwen-Flash: ~$0.05-0.40/day (cheapest)
• Qwen-Plus: ~$0.10-0.60/day (default)
• Qwen-Max: ~$0.15-0.80/day (best quality)

Cost Factors: Query frequency, market volatility, Reasoning Effort level, and timeframe selection all affect daily costs. Use Max Daily Cost setting to control spending.

For Beginners:

• GPT-5.2 Instant (default) - reliable, fast, and cost-effective
• Gemini 2.5 Flash - excellent value with good performance
• DeepSeek-Chat or Qwen-Plus - ultra-affordable with solid performance
• Grok-3 Mini - balanced budget option

For Best Performance:

• GPT-5.2 Thinking (with High effort) - enhanced 'xhigh' reasoning
• Claude Opus 4.5 - deep reasoning with extended thinking
• Gemini 3 Pro - latest from Google
• Grok-4 - latest from xAI
• Qwen-Max - best quality from Alibaba

For Complex Market Analysis:

• GPT-5.2 Thinking with High effort (xhigh reasoning mode)
• Claude Opus 4.5 (extended thinking mode)
• DeepSeek-Reasoner (affordable reasoning mode)

For Testing/Budget:

• DeepSeek-Chat (ultra-cheap & solid performance)
• Qwen-Flash (cheapest with free tier)
• Gemini 2.5 Flash-Lite (fastest & cheap)
• Claude Haiku 4.5 (fast & economical)
• GPT-5 Nano (smallest OpenAI model)

No! You only need ONE API key from the provider you choose to use.

The EA supports 6 providers to give you flexibility:

• Start with one provider (e.g., OpenAI, Gemini, or DeepSeek)
• Later, you can add other providers to test different AI approaches
• Switch providers anytime by changing the AI Model Selection
• Only the selected provider's API will be charged
• Having multiple keys allows you to compare performance or switch if one provider has issues

For Fallback: If you configure a fallback model from a different provider, you'll need API keys for both providers (primary and fallback).

Tip: Many traders start with GPT-5.2 Instant (default), Gemini (free tier available), or DeepSeek (ultra-affordable), then experiment with others based on their trading needs.

Yes! The EA includes a fallback strategy that works without AI. You can also:

• Use Strategy Tester (AI disabled automatically)
• Set very long query intervals to minimize costs
• Use Conservative autonomy level for suggestions only

Check these common issues in order:

1. WebRequest URLs: ⚠️ MOST COMMON - Verify the 6 URLs are allowed in MT5 Options (Step 2.5). Without these, EA cannot connect to AI providers!
2. API Key: Ensure the correct key for your selected provider is entered and has credits/balance
3. Provider Match: Make sure your selected model matches the API key entered (e.g., GPT-5.2 needs OpenAI key, not Claude key)
4. Confidence Level: Lower the minimum confidence if too high (try 50% for testing)
5. Time Filter: Check if current time is within allowed hours (or disable time filter temporarily)
6. Autonomy Level: If set to Conservative, AI only suggests but doesn't execute - you must trade manually
7. Journal Tab: Check for error messages like "URL not allowed" or "invalid API key"
8. Internet Connection: Verify MT5 has internet access and no firewall is blocking connections

While possible, we recommend:

• Use different Magic Numbers for each EA
• Stagger query times to avoid API rate limits
• Monitor combined API costs carefully
• Consider using different AI models for variety
• Ensure adequate margin for all EAs combined

The Fallback Provider is a backup AI that activates when your primary provider experiences errors (server outages, rate limits, timeouts).

How it works:

• After 2 consecutive server errors from primary, EA switches to fallback
• After 1 successful fallback query, EA tests primary again
• If primary recovers, EA automatically switches back
• GUI shows: ⚡ model-name [FALLBACK] when using backup

Recommended Combinations:

• Primary: Gemini 2.5 Flash → Fallback: Claude Sonnet 4.5
• Primary: GPT-5.2 Instant → Fallback: DeepSeek-Chat
• Primary: Qwen-Plus → Fallback: Gemini 2.5 Flash

Important: Always use different providers for primary and fallback. If both use the same provider and their servers are down, both will fail.

In MT5, look for these indicators:

• 🔵 Blue hat icon in top-right = EA running
• 🔴 Red/Gray hat = EA stopped or error
• Check the Experts tab for AI query logs
• Look for "AI Decision" messages in Journal