Docs/API Reference/Models

Models

Choose the model that fits your use case. Smart is great for quick analysis, while Deep Research models provide comprehensive multi-agent analysis with specialized agents.

#Smart

Skills-based agent for fast, efficient analysis. Discovers and uses financial skills dynamically via HTTP and sandboxed Python.

alphagent-smart

Recommended for most use cases

Skills-based analysis with code execution and web search. Great for market analysis, earnings summaries, and general queries.

#Deep Research Models

Multi-agent models with 5 specialist agents working in parallel, each with domain-filtered skills. Best for comprehensive analysis.

alphagent-deep-research-fast

5-agent • Speed optimized

Multi-agent research with faster response time. A Router delegates to 5 specialist agents with domain-specific skills.

alphagent-deep-research-pro

5-agent • Highest quality

Multi-agent research with pro models. Best for investment research, due diligence, and comprehensive market analysis.

#Agent Architecture

Deep Research models use a Router that delegates to 5 specialist agents, each with domain-filtered skills plus a Verifier:

Fundamental Analyst

DCF valuation, financial health, peer comparison

Portfolio Manager

Portfolio risk, correlation analysis, macro overlays

Derivatives Trader

Options Greeks, stress testing, P&L attribution

Technical Analyst

Price action, liquidity, corporate actions

Sentiment Analyst

Reddit sentiment, macro/economic indicators

Verifier

Validates claims against sources

#Model Comparison

ModelTypeBest For
alphagent-smartSkills-basedGeneral queries, most use cases
alphagent-deep-research-fast5-agentComprehensive research, faster
alphagent-deep-research-pro5-agentDeep due diligence, highest quality