Understanding Request Sizes

Request size directly affects analysis speed and cost. Larger requests take longer to process and cost more (for cloud providers). Understanding how to manage request sizes helps optimize performance.

Size Management Strategies

Use STRICT Diff Scope

STRICT scope includes only changed lines, significantly reducing request size while maintaining analysis quality for most cases.

Selective Full Content

Only include full file content when necessary. Diffs alone are often sufficient and much smaller.

Limit Related Context

Configure related context caps appropriately. More context helps but increases size—find the right balance.

Intelligent Batching

The plugin automatically batches large commits, but you can optimize this:

  • Keep commits focused (fewer files = faster analysis)
  • Trust the batching system to handle large commits
  • Review results from all batches

Provider Selection

For Speed

Cloud providers are generally faster than local options. Use them when speed is important.

For Cost

Local providers have no API costs. Use them when cost is a concern and speed is acceptable.

Model Selection

Standard Models

Use standard models (not Think Harder) for routine work. They're faster and sufficient for most cases.

Think Harder Selectively

Reserve Think Harder mode for important changes where the extra depth is valuable.

Conclusion

Performance optimization requires balancing analysis quality with speed and cost. By managing request sizes, using appropriate models, and trusting the batching system, you can get fast, efficient analysis.

Ready to optimize? Install AI Diff Review and configure settings for optimal performance.