Changelog

GPT-5.1 is now available in Augment, served through OpenAI's Priority Processing for faster, more reliable responses.

Why Priority Processing?

We don't think you should settle for best-effort latency when you're shipping production code. Priority Processing delivers faster, more consistent token generation with higher uptime guarantees, even during peak demand.

What improved:

• 15% lower latency compared to GPT-5
• Closer alignment with human-authored pull requests
• 72.5% file overlap with ground truth (vs. 65% for Sonnet 4.5 and 66% for GPT-5)
• More efficient reasoning: works longer while spending fewer tokens per action

This translates to noticeably faster performance on straightforward tasks, with better focus on complex multi-file edits.

Credit costs: Despite Priority Processing being more expensive per token, GPT-5.1's improved efficiency means you'll see similar credit costs to Sonnet 4.5 for comparable tasks. The model does more with less.

We optimized for quality and reliability. Try it and let us know what you think.