Agentbase offers three distinct agent modes that optimize performance and cost through intelligent model selection. Rather than using a single model, each mode employs a mixture of models including custom models specifically designed to reduce agent execution steps and overall costs.

​
Why Model Mixtures?

Agent costs are primarily driven by input tokens (approximately 80% of total cost), as agents must send the same context repeatedly across multiple steps. By using intelligent model mixtures, we can:
  • Reduce execution steps through smarter planning
  • Lower input token costs by completing tasks more efficiently
  • Optimize for specific use cases with specialized model combinations

​
Available Modes

Fast Mode

Most cost optimized and efficient

Flash Mode

Thinking capability with good speed

Max Mode

Highest capability for complex tasks

​
Ready to Try AgentBase?

Experience the power of intelligent agent modes with your own tasks. Start with Fast mode for cost-effective automation, then explore Flash and Max modes as your needs grow.

Sign Up & Get Started

Get free credits with your work email and test all three modes today

​
What’s Next?

Python SDK

Use Agentbase with Python

TypeScript SDK

Use Agentbase with TypeScript/JavaScript
Mode selection is automatically optimized based on task complexity, but you can override the default behavior by specifying your preferred mode.