HaiAI123

Curated Global AI Tools Directory

Industry update
Flat

Anthropic makes Claude Opus 5.5 the default model on paid plans

11/15/2026 — 11/15, 23:17·2 sources (all)·4 reports (all)

Key takeaways

  • Anthropic has made Claude Opus 5.5 the default model on all paid plans and published guidance telling users to drop older prompting habits,…
  • in internal tests Opus 5.5 at Medium matched the previous Opus 5 at High.

Background & analysis

From October 30 to November 8, 2026, a developer ran 209 calls against Claude Opus (claude-opus-5-5) over 10 days, spanning 22 kinds of tasks. The total cost came to $23.7465, or $0.11362 per call, with a median call time of 18.1 seconds. In the same window, Claude Sonnet (claude-sonnet-5-5) was called 156 times at $0.0165 per call, with a median of 2.3 seconds and a slowest call of 19.2 seconds, meaning Opus cost roughly 7 times as much per call and its median run was about 8 times longer. That benchmark was published on 2026-11-12 at 18:00.

On 2026-11-15, The Decoder and DEV Community · AI reported in turn that Anthropic has made Claude Opus 5.5 the default model in Claude on all paid plans, covering Pro, Max, Ultra, and Team, and has published a guide for knowledge work that asks users to drop prompting habits inherited from earlier model generations. Anthropic states that the model's thinking is always on, so instructions such as "think hard" or "think carefully" are unnecessary and can be deleted, while users should start at Medium effort and let the model adjust its own reasoning depth. The company says internal testing showed Opus 5.5 at the Medium setting matching the High setting of the previous Opus 5, and that removing those instructions made replies start sooner with no clear drop in quality. The guide also lists seven prompting habits worth revisiting and recommends treating Opus 5.5 as the daily driver for research, analysis, and finished documents, decks, and spreadsheets. The episode currently stands at the point where the guide has been released and the default-model change is in effect.

AI-generated from 4 reports · updated 2 days ago

Latest turnThe author points to Anthropic's November 12 guide to using Claude Opus 5.5 for knowledge work, which lists seven prompting habits worth revisiting. In Anthropic's tests, dropping instructions like "think hard" or "think carefully" made replies start sooner with no clear quality drop, and the author ran the chart-to-report example in Claude himself.

Related AI tools

updream
24-hour heatHeat index 34 · peak 102 · 23h ago
24 hours agonow

Reports on this story headlines open the original

Nov 15
  1. The author points to Anthropic's November 12 guide to using Claude Opus 5.5 for knowledge work, which lists seven prompting habits worth revisiting. In Anthropic's tests, dropping instructions like "think hard" or "think carefully" made replies start sooner with no clear quality drop, and the author ran the chart-to-report example in Claude himself.

    DEV Community · AIAI score 57
  2. Anthropic published a knowledge-work guide that positions Claude Opus 5.5 as the default model across Pro, Max, Ultra and Team plans and asks users to drop prompting habits from older generations. The guide says thinking is always on in Opus 5.5, so directives like "think hard" can be removed, and recommends starting at the Medium effort level.

    DEV Community · AIAI score 64
  3. Anthropic has made Claude Opus 5.5 the default model on all paid plans and published guidance telling users to drop older prompting habits. The company says phrases like "think hard" no longer help, and that starting at the Medium effort level matches Opus 5 on High in internal tests, according to The Decoder.

    The DecoderAI score 63
Nov 12
  1. A developer ran 209 Claude Opus (claude-opus-5-5) calls over 10 days across 22 job types, spending $23.7465 total — about $0.11362 per call, with a median latency of 18.1 seconds. In the same window, Claude Sonnet (claude-sonnet-5-5) handled 156 calls at $0.0165 each, a 2.3-second median and a 19.2-second worst case. Opus cost roughly 7x Sonnet per call and ran about 8x longer at the median.

    DEV Community · AIAI score 54

Other stories people are talking about

How is heat calculated?About the method

Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.

This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.

Surge
Discussion rising fast
New
First report within 6 hours
Rising
Still gathering discussion

Back to AI industry updates →