We’re excited to announce Composer 1.5, our latest agentic coding model that significantly improves on Composer 1.
Key features:
- Adaptive thinking: Reasons through your codebase and adjusts thinking time based on problem difficulty—staying fast for simple tasks while diving deeper on complex ones
- Self-summarization: Handles longer tasks by intelligently summarizing context when needed, maintaining accuracy across varying context lengths
- Stronger performance: Substantial improvements on our internal benchmark of real-world coding problems, especially on challenging tasks
Composer 1.5 balances speed and intelligence for daily use, and we recommend it for interactive coding workflows.
It’s available now!
Check out the pricing details and technical blog post to learn more about the training approach and capabilities.
We’d also love for you to share your first impressions here!