🔧 CREATE AGENT Pricing

Build custom LLRM-SQL agents with natural language descriptions

How CREATE AGENT Works

CREATE AGENT allows authorized users to build custom agents by providing a natural language description, agent name, and LLRM-SQL definition. The system processes your input through a 5-stage pipeline including prompt analysis, SQL generation, validation, and repository storage with CLIP-like similarity indexing.

🔐 Access Control & Authorization

Only ADMIN and ENTERPRISE users can create agents. This feature is not available to PRO or FREE tier users.

ADMIN Users: Full access with no monthly limits
ENTERPRISE Users: Full access within monthly spending limits
PRO Users: ❌ Cannot create agents
FREE Users: ❌ Cannot create agents

👑 ADMIN Tier

Unlimited agent creation with full system access

No monthly spending limits
Priority processing
Advanced debugging tools
System-wide agent management

Agent Complexity	Cost per Creation
Simple Agent	$0.007
Medium Agent	$0.014
Complex Agent	$0.035
Very Complex Agent	$0.070

🏢 ENTERPRISE Tier

Professional agent creation within spending limits

Monthly spending limits apply
Standard processing priority
Team collaboration features
Usage analytics

Agent Complexity	Cost per Creation
Simple Agent	$0.007
Medium Agent	$0.014
Complex Agent	$0.035
Very Complex Agent	$0.070

💻 Cost Breakdown

CREATE AGENT costs are based on AWS infrastructure usage with 50/50 split:

Component	AWS Cost	SmartMLChain Cost	Total Cost
Lambda Processing	$0.001	$0.001	$0.002
S3 Storage	$0.0005	$0.0005	$0.001
Embedding Generation	$0.002	$0.002	$0.004
Base Total	$0.0035	$0.0035	$0.007

Additional costs may apply for complex LLRM-SQL validation and GPU processing.

💡 Cost Examples

Example 1: Simple Stock Tracker Agent

Description: "Track daily stock prices for a given symbol"

LLRM-SQL: Basic SELECT with stock_price() function

Processing: Standard validation, simple embedding

Total: $0.007

Breakdown: $0.0035 AWS + $0.0035 SmartMLChain

Example 2: Multi-Source Analysis Agent

Description: "Analyze market trends using multiple data sources and generate PDF reports"

LLRM-SQL: Complex CTE with multiple joins, AI model calls, PDF generation

Processing: Advanced validation, complex embedding, GPU usage

Total: $0.035

Breakdown: $0.0175 AWS + $0.0175 SmartMLChain

Example 3: Enterprise Dashboard Agent

Description: "Real-time executive dashboard with predictive analytics and automated alerts"

LLRM-SQL: Very complex with ML models, real-time data, email integration

Processing: Extensive validation, high-dimensional embedding, GPU processing

Total: $0.070

Breakdown: $0.035 AWS + $0.035 SmartMLChain

⚡ Cost Optimization Tips

Simplify LLRM-SQL: Use fewer CTEs and simpler logic where possible
Optimize Descriptions: Clear, concise descriptions reduce embedding costs
Reuse Patterns: Build on existing agent patterns rather than creating from scratch
Batch Creation: Create multiple similar agents in one session to reduce overhead
Test Locally: Validate LLRM-SQL syntax before submitting to reduce failed attempts
Use Templates: Start with proven agent templates to minimize complexity
Monitor Usage: Track which agents are actually used to avoid creating unnecessary ones

🔄 Agent Creation Pipeline Costs

Pipeline Stage	Processing	Cost Impact
Stage 0: Matching Engine	Similarity search	Low
Stage 1: Prompt Analyzer	NLP processing	Medium
Stage 2: SQL Generator	Code generation	Medium
Stage 3: Validator	Syntax validation	High
Stage 4: Repository	Storage & indexing	Low