Prompt Library ⚙️ Technical Long-Context Document Analysis
GPT-4o ⚙️ Technical Basic

Long-Context Document Analysis

Use million-token context AI models to analyze entire documents, codebases, or research collections and extract structured insights in one pass.

👁 0 views ⎘ 0 copies ♥ 0 likes

The Prompt

# Long-Context Document Analysis

You are an AI systems specialist. Design a long-context analysis workflow for [DOCUMENT SET, e.g. a 500-page legal filing, an entire codebase, a year of earnings call transcripts].

## Context Setup

- Model to use: [MODEL, e.g. Claude Fable 5.1, GPT-6 Astra, Gemini 3.8 Flash]
- Context window available: [TOKENS]
- Document size: [APPROXIMATE SIZE]

## Analysis Goal

What are you trying to extract or answer? [ANALYSIS OBJECTIVE, e.g. identify all risk clauses, find all references to a specific topic, summarize the key arguments]

## Prompt Structure

Design the system prompt and user prompt for this analysis. Include instructions for:
- How to handle ambiguous sections
- Output format (JSON / markdown / plain prose)
- How to flag low-confidence findings

## Chunking Strategy (if needed)

If the document exceeds the context window, describe how to split, process, and merge chunk outputs.

## Verification Step

How do you validate that the model has not missed sections or hallucinated references?

📝 Fill in the blanks

Replace these placeholders with your own content:

[DOCUMENT SET, e.g. a 500-page legal filing, an entire codebase, a year of earnings call transcripts]
[MODEL, e.g. Claude Fable 5.1, GPT-6 Astra, Gemini 3.8 Flash]
[TOKENS]
[APPROXIMATE SIZE]
[ANALYSIS OBJECTIVE, e.g. identify all risk clauses, find all references to a specific topic, summarize the key arguments]

How to use this prompt

1
Copy the prompt

Click "Copy Prompt" above to copy the full prompt text to your clipboard.

2
Replace the placeholders

Swap out anything in [BRACKETS] with your specific details.

3
Paste into GPT-4o

Open your preferred AI assistant and paste the prompt to get started.