Key Takeaways

  • Anthropic fixed the chatty verbosity of Opus 5, but Opus 5.5 stops narrating its actions, which makes execution feel artificially slow.
  • GPT-6 Sol costs about half as much as Claude Opus 5.5 while delivering higher response speeds and tighter token counts.
  • Claire Vo abandoned Claude for months in favor of OpenAI tools like Sol and Astra for everyday knowledge work.
  • Opus 5.5 holds the advantage in complex B2B frontend code, but Sol wins on writing clarity, email triage, and raw cost efficiency.

The Problem With Mute Models

For months, Claire Vo stopped opening Anthropic tools altogether. As she put it, “I truly have not opened Claude in months for day-to-day tasks. I've just been like really really into ChatGPT, really into Codex, really into Astra, really into Sol.”

Opus 5 had developed a reputation for being irritatingly verbose. It generated walls of explanatory filler before doing what you asked. Anthropic listened to user frustration and overcorrected with Opus 5.5.

Vo tested the new release across real workloads like PRDs, email triage, frontend coding, and video editing. Her immediate reaction was relief: “The feedback I gave to the Anthropic team is like not annoying. I found myself annoyed zero times with Opus 5.5.”

Yet that silence introduced a new ergonomic problem. “I think part of the way they made Opus 5.5 not annoying is they had it shut up,” Vo explained. “I also just found that it wouldn't narrate its work.” When a model works through a multi-step task without streaming its thought process or status updates, the user stares at a blank screen. That silence destroys perceived speed.

The 2x Price Gap and Everyday Speed

Perceived speed matters, but raw latency and pricing decide where teams route API calls. In head-to-head testing against GPT-6 Sol, Claude lagged behind.

“The other thing is Opus 5.5 is pretty fast, but not as fast as Sol is,” Vo noted. “So I was able to test both and I constantly felt like the GPT models were way faster.”

Then comes the bill. Anthropic priced Opus 5.5 at roughly double the cost of GPT-6 Sol. When Vo ran her initial blind benchmarks, that pricing structure had not been published yet. Once the numbers dropped, the math shifted.

"Opus 5.5 is twice as expensive as GPT-6 Sol," Vo said. “And the pricing wasn't out when I tested these. And this is really going to impact how I think about using these models.”

If you run automated evaluation loops, customer-facing agents, or high-volume summarization, paying double for a model that runs slower is a bad trade. Sol delivers concise bullet points, rapid token generation, and clean text outputs for half the price. Opus 5.5 earns its keep when you need deep frontend logic or complex component architecture, but using it as a default text engine wastes budget.

What to Do With This

Audit your team's API routing table by task complexity this week. Direct all high-volume text operations, PRD drafts, and email triage pipelines to GPT-6 Sol to cut inference costs by 50%. Reserve Claude Opus 5.5 strictly for multi-file frontend code generation and complex agentic tasks where Anthropic still holds the quality edge.