Skip to main content
PromptQuorumBuilt for humans. Structured for AI.
Medium complexityUsed to win Singapore's GPT-4 competition

CO-STAR Framework

Context · Objective · Style · Tone · Audience · Response

Credited to GovTech Singapore's Data Science & AI team, and used to win Singapore's first GPT-4 prompt engineering competition. Maximum control over voice, audience, and output format.

By · Last updated

CO-STAR framework fields overview
Definition
The CO-STAR framework (Context · Objective · Style · Tone · Audience · Response) is a prompt engineering structure that breaks your AI request into 6 discrete fields. It is best suited for business communication and marketing copy.

The 6 Fields

1

Context

Background information the AI needs to understand the task and situation.

2

Objective

The specific goal or outcome you want the AI to achieve.

3

Style

The writing style — e.g., formal, conversational, journalistic, academic.

4

Tone

The emotional register — e.g., persuasive, empathetic, authoritative, friendly.

5

Audience

Who the output is for. Shapes vocabulary, depth, and assumptions.

6

Response

The exact output format — e.g., bullet list, 500-word article, JSON, table.

Real Example

Scenario: Writing a product launch email for a SaaS tool targeting CTOs

CO-STAR Prompt

Context: PromptQuorum is a multi-model AI tool launching in July 2026. Objective: Write a product launch email that drives signups. Style: Clear and direct, no fluff. Tone: Confident and professional. Audience: CTOs at mid-size tech companies. Response: 200-word email with subject line.

When to Use CO-STAR

When to use CO-STAR
Best for
  • ✓Business communication and marketing copy
  • ✓Content where voice and tone matter (brand writing)
  • ✓Professional writing for a specific audience
  • ✓Any task where you need to control style separately from tone
Not ideal for
  • ✗Quick one-off queries where structure adds no value
  • ✗Sequential multi-step tasks (use RISEN instead)
  • ✗Technical tasks where examples matter more (use TRACE)

Frequently Asked Questions

What does CO-STAR stand for?

CO-STAR stands for Context, Objective, Style, Tone, Audience, and Response — six fields that give you precise control over AI output.

Where does CO-STAR come from, and did it win the Singapore GPT-4 competition?

CO-STAR is credited to GovTech Singapore's Data Science & AI team. Sheila Teo used it to win GovTech's first GPT-4 prompt engineering competition, which drew over 400 participants, and described the approach in a Towards Data Science article. The framework did not win by itself; the win was Teo's. Its core design idea is separating Style (presentation) from Tone (emotion).

When should I use CO-STAR vs CRAFT?

Use CO-STAR when you need fine-grained control over voice and audience. Use CRAFT when you want a Role-based framing with a clear Target persona. Learn how to measure the effectiveness of your CO-STAR prompts at https://www.promptquorum.com/prompt-engineering/how-to-evaluate-prompt-quality.

Sources

Related Frameworks

Use CO-STAR inside PromptQuorum

All 9 frameworks are built in. Write, optimize with your own LLM, dispatch to 25+ AI models, and run consensus analysis — free, no account required.

Download the Beta
MCP