Sunday, 15 November 2026, story 1 of 7
Anthropic tells Opus 5.5 users to drop old prompting habits
The Decoder reports Anthropic says Medium effort on Opus 5.5 matches Opus 5 on High, according to the company's internal tests.

What happened
Anthropic is advising users of Claude Opus 5.5 to drop prompting habits from older models, The Decoder reports. The model is now the default on Pro, Max, Ultra and Team plans, according to Anthropic. The company says phrases such as "think hard" are unnecessary because the model reasons on its own and scales depth to the task and the chosen effort level. In its tests, replies without them started faster with no measurable quality drop.
Anthropic also recommends starting at Medium effort, the default. It says Opus 5.5 on Medium matched Opus 5 on High in internal tests. These are the company's own figures.
What it means
Prompt templates that carry reasoning nudges from earlier model generations are now worth auditing. If Anthropic's tests hold, those phrases add clutter and no quality. The effort setting becomes the main control, and Anthropic's advice to move up from Medium only when needed points to a cost lever.
The Sonnet argument deserves a check rather than trust. Anthropic says fewer steps can partly offset Opus's higher per-token price, and that Sonnet at Extra High or Max costs about as much. Both claims come from the company, and the source gives no prices or test details. It also describes the setting in the model menu next to the send button, so it is unclear how the guidance applies outside that interface. Independent comparisons on real workloads would show how far the claims carry.
What to do: Cut think-hard phrases from prompts, run a few tasks at Medium, and compare cost and quality against your current Sonnet or High settings.
Source
Filed under Models. Published .
"What it means" is analysis, not reporting. Found an error? Write to daily@vibitely.com.