# A Lower-Cost AI Matched a Frontier Model in Half as Many Iterations by Gabriel Grinberg > "Across 118 real app builds, Claude Sonnet 5.5 produced apps that scored level with Opus 5. It got there in 3.6 iterations per build on average, where Opus 5 took 7.7. It had the fewest failed tool calls of any model we compared." > **— Gabriel Grinberg**, *Introducing Claude Sonnet 5.5, September 28, 2026* ## Sources and Context - **Direct publication:** [Introducing Claude Sonnet 5.5](https://www.anthropic.com/claude-sonnet-5-5) — published September 28, 2026. - **Exact published statement:** “Across 118 real app builds, Claude Sonnet 5.5 produced apps that scored level with Opus 5. It got there in 3.6 iterations per build on average, where Opus 5 took 7.7. It had the fewest failed tool calls of any model we compared.” - **Context and evidence boundary:** Grinberg described Base44's evaluation across 118 real application builds, comparing a lower-cost model with Anthropic's frontier tier. - **Weekly source dossier:** [[research/Weekly Reminder Hunt - 2026-10-02|Weekly Reminder Hunt — October 2, 2026]] preserves the selection rationale and the other authoritative statements reviewed in the same production cycle. ## Related Articles and Collections - **Collection:** [[collections/Simple Reminders|Simple Reminders]] - **Collection:** [[collections/Machine Succession|Machine Succession]] - **Article:** [[articles/The Gigawatt Is the New Measure of War|The Gigawatt Is the New Measure of War]] ## Related Topics - [[wiki/Gabriel Grinberg|Gabriel Grinberg]] - [[wiki/AI Distribution|AI Distribution]] - [[wiki/AI Price-Performance|AI Price-Performance]] - [[wiki/AI Access|AI Access]] ## Share on Social Media ``` "Across 118 real app builds, Claude Sonnet 5.5 produced apps that scored level with Opus 5. It got there in 3.6 iterations per build on average, where Opus 5 took 7.7. It had the fewest failed tool calls of any model we compared." — Gabriel Grinberg, Introducing Claude Sonnet 5.5, September 28, 2026 https://bryantmcgill.com/simple-reminders-grinberg-lower-cost-ai-matched-frontier-model-in ```