How I AI
How I AI

Sep 22, 2026 · 39 min

Opus 5.5 wins breadth while GPT-6 wins the vibe

Opus 5.5 vs. GPT-6 Sol: which model won my blind taste test?

The blind test separates personal preference from dependable performance across creative work, coding, research, and long-running agents.

3 key takeaways
  1. 1GPT-6 Astra wins Claire’s personal preference, while GPT-6 Sol stands out for clear writing and value.
  2. 2Opus 5.5 performs best across the broadest task range, especially agentic and long-running work.
  3. 3Creative generation remains uneven across models, with video editing weak and Barbie anatomy still visibly flawed.

Don't miss

The blind-test reveal separates Claire’s favorite, GPT-6 Astra, from the model that performs best across the full benchmark, Opus 5.5.

The brief

Claire Vo records a live blind comparison of Opus 5.5, GPT-6 Astra, GPT-6 Sol, and GPT-6 Luna across a redesigned benchmark spanning work and creative tasks.

The test moves from inbox triage and email drafting to frontend prototypes, backend coding, research, and computer use, judging both output quality and interaction style.

Opus 5.5 feels more usable than Claire expected, with cleaner communication and stronger agentic performance, even as its conservative personality remains a drawback.

The creative tests expose a wider weakness: SVG quality varies sharply, video editing is poor across the board, and generated Barbie still struggles with anatomy.

GPT-6 Astra wins Claire’s personal preference and Sol earns praise for writing and value, but Opus 5.5 takes the broadest overall performance win.

Weekly email

State of the Pod, every Monday

The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.

Listen to the full episode and explore every guest, topic, and moment on PodLume.