August
Getting Started

Choosing the Right Mode

August's Assistant offers multiple operating modes, each optimized for a different point on the speed-versus-depth curve. Select the mode that matches your task's priorities.

Mode selector

Mode Comparison

Each mode positions differently on speed and depth. Choose based on whether you need fast results, visible reasoning, or precise control.

Mode

Speed

Depth

Best For

Agent (default)

Fast

Multi-step reasoning + tool use

Daily tasks, drafting, research

Genius

Slow

Pattern recognition & intuition

Fluent, high-level output where polish matters

Deep Thinking

Moderate

Methodical, step-by-step reasoning

Problems that benefit from visible chain-of-analysis

Manual

User-controlled

User-directed intent control

Toggle Create / Research / Analyze for precise output direction

Agent Mode

Agent is the default for most work. It balances speed with multi-step reasoning and tool use, making it suitable for day-to-day tasks like drafting, research, and document analysis.

Use Agent when you need a fast, capable response and don't require visibility into the reasoning process.

Agent mode currently uses GPT-5.5 (low reasoning effort) as the default model for planning, tool execution, and response generation. For structured outputs such as tables and formatted extracts, August prefers GPT-5.5 with GPT-5.1 as a fallback.

Genius Mode

Genius Mode is slow and thoughtful, designed for multi-step assembly and polished, high-level output. It relies on pattern recognition and intuition to produce cohesive, well-structured results without exposing its internal reasoning.

Use Genius Mode when:

  • You're producing formatted outputs in Excel, PowerPoint, Word, or PDF

  • You need multi-step assembly across multiple documents

  • Polish, voice, and document-ready quality matter more than speed

For detailed guidance, see Genius Mode Overview.

Deep Thinking

Deep Thinking works more slowly and methodically. It explicitly breaks problems into steps, tests assumptions, and surfaces its chain of analysis before giving conclusions.

Use Deep Thinking when:

  • You need to see the reasoning, not just the result

  • The problem benefits from visible chain-of-analysis

  • You're working through dense legal analysis or hard reasoning tasks

Manual Mode

Manual gives you tighter control over how the Assistant behaves during the task. Toggle between Create, Research, and Analyze to direct the output precisely.

Use Manual when you already know the exact workflow you want to drive step by step and need precise intent control.

How to Choose

  • Start with Agent for most day-to-day work.

  • Switch to Genius for multi-step assembly work across multiple documents.

  • Move to Deep Thinking for denser analysis or harder reasoning tasks where you want to see the reasoning chain.

  • Use Manual when you want to guide each step more directly with precise intent control.

No matter which mode you use, validate the output against cited source material before relying on it. Every output traces back to its source documents through citations.

Was this helpful?