signalLenny's Newsletter2026-09-23
I left Claude for months. Opus 5.5 is why I'm back
In this podcast episode, Claire Vo explains why she returned to Claude after months away, citing Opus 5.5's 40% cost reduction, faster speed, and new alignment approach. After testing it on agentic tasks, frontend prototyping, and SVG benchmarks, she praises its frontend capabilities but notes persistent annoyances and an unexpected refusal capability. She also details where Codex still outshines Opus 5.5 and how she now splits her model stack.
- for who
- AI developers and users of agentic tools evaluating Claude models
- why now
- Opus 5.5 just shipped - 40% cheaper, new alignment.
- what changes
- Prompts a shift toward Opus 5.5 for prototyping and agentic workflows
- to do
- Benchmark Opus 5.5 on your own agentic tasks and compare cost
key points
- Opus 5.5 is 40% cheaper and faster than Opus 5
- Four agentic tasks, a ChatPRD redesign, and an SVG benchmark tested
- Firm refusal and frontend prototyping stand out as strengths
#Claude Opus 5.5#model evaluation#ai tools5 sources · confidence high
score
score 8 out of 10. 0-10: how dense the facts are, multiplied by how much you can do with them after reading. 8+ means the topic's evidence bar is met: benchmarks and availability for a new model, amount and investors for a funding round, revenue figures for a solo-money story. Below 5 an item does not enter the digest. A press release scores 3 or less, a reprint loses 2, anything older than 14 days loses 1, a headline that misleads loses 3.
read the source