Canvas Settings

System Prompt Configuration

Customize the system prompt that guides the AI behavior in Canvas. This prompt defines:

Response format and structure
Chart and table generation rules
Database query capabilities
Data visualization guidelines

⚠️ Warning: Modifying the system prompt can significantly change the AI behavior. Make sure to maintain the JSON response structure with "explanation" and "artifacts" fields for proper functionality.

System Prompt Editor

0 characters

Preview

Chat History & Context Settings

Configure how the AI manages conversation context and memory across sessions.

💡 Tip: These settings apply to new chat sessions. Current sessions use their existing settings.

AI Model

Context window: 200,000 tokens

Context Window Size (tokens)

80% of model's capacity

Maximum Messages in Context

How many recent messages to keep in context

Auto-Summarize Old Messages

Enable smart summarization when context limit is reached

Preserves conversation context while staying within token limits

Model Pricing (per 1K tokens)

Minimax M2

Input: $0.0002 | Output: $0.0006

Claude 3.5 Sonnet

Input: $0.0030 | Output: $0.0150

Claude 3.5 Haiku

Input: $0.0010 | Output: $0.0050

GPT-4o

Input: $0.0025 | Output: $0.0100

GPT-4o Mini

Input: $0.0001 | Output: $0.0006

Gemini Pro

Input: $0.0005 | Output: $0.0015