Updated

A sculpture of folded terracotta clay and pleated ivory paper on a dark background, abstract hero artwork for Claude Opus 5.5

Anthropic

Claude Opus 5.5

Anthropic's newest Opus, near Fable 5.1 on most work and built for long agentic coding.

Chat with Claude Opus 5.5 in Super AgentView pricing

What is Claude Opus 5.5?

Claude Opus 5.5 is Anthropic's Opus model for long-running agentic coding and knowledge work, released on September 22, 2026 as the successor to Claude Opus 5.

Anthropic says it performs at the level of Claude Fable 5.1 on most work while costing about 40% less to run than Opus 5 at default settings, and that it generates output more than 30% faster than Opus 5. It writes more clearly, puts the most important information first, and is less likely than Opus 5 to take hard-to-reverse actions. Anthropic's docs recommend starting with Opus 5.5 for most workloads.

Provider
Anthropic
Context window
1M tokens
Input
Text, Image
Reasoning
Low to max
Max output
128K tokens
Knowledge cutoff
June 2026
Released
Sep 22, 2026

How Claude Opus 5.5 performs

Published by Anthropic

Results Anthropic reported for Claude Opus 5.5. Each lab runs its own setup, so compare these with other models only inside the same table.

Benchmark results Anthropic published for Claude Opus 5.5
BenchmarkScoreSetting
Terminal-Bench 4.066.4%extra high effort
FrontierCode v1.1 (Main)54.4%max effort
CursorBench 4.057.8%max effort
Humanity's Last Exam67.7%with tools, max effort
Terminal-Bench-Science 0.158.7%max effort
OSWorld 2.181.8%partial, max effort
Chartography89.0%with tools, max effort

Source: anthropic.com

What you can do with Claude Opus 5.5

  • Close to Claude Fable 5.1 on most work, according to Anthropic

  • Output more than 30% faster than Opus 5

  • Puts the most important information first when it writes

  • Best score of any model so far on Anthropic's automated behavioral audit

  • More resistant to prompt injection than Opus 5

How to prompt Claude Opus 5.5

  • Give it the whole project goal; Anthropic built it for long agentic work.
  • Super Agent starts it at medium effort; drop to low for quick edits, raise it for big plans.
  • Ask for the conclusion first. Anthropic says it already tends to lead with the main point.

Where Claude Opus 5.5 fits best

  • Long, multi-step creative projects

  • Big migrations and refactors

  • Reports that lead with the answer

  • Agent runs where a wrong action is costly

How Claude Opus 5.5 compares

Claude Opus 5.5 compared with other chat models on Flashloop
ModelLMArena textLMArena codingContext windowReasoning
Claude Opus 5.51509#21548#41M tokensLow to max
Claude Fable 51503#41551#21M tokensLow to max
Claude Fable 5.11503#51538#131M tokensLow to max
Claude Opus 51491#131528#231M tokensOff to max
Gemini 3.7 Flash1490#161520#361M tokensLow, medium, high

Good to know before you start

  • No thinking-off mode; quick edits run at low effort
  • Anthropic's docs rate its latency moderate, against fast for Claude Sonnet 5.5
  • Text output only; Super Agent sends visual work to Flashloop's generation models

Start chatting with Claude Opus 5.5

Chat with Claude Opus 5.5 in Super Agent

  1. Select Claude Opus 5.5

    Open Super Agent on Flashloop and choose Claude Opus 5.5 in the model picker.

  2. Enter your prompt

    Ask for what you need: an answer, a plan, or the images and videos you want the agent to make.

  3. Send and keep going

    Send the message. Claude Opus 5.5 answers in Super Agent and runs any generations you asked for.

Questions about Claude Opus 5.5

What is new in Claude Opus 5.5 compared with Opus 5?

Anthropic says Opus 5.5 performs close to Fable 5.1 on most work, costs about 40% less to run than Opus 5 at default settings and generates output more than 30% faster. It also has the best score of any model on Anthropic's automated behavioral audit and resists prompt injection better than Opus 5.

What is Claude Opus 5.5 best at?

Long-running agentic coding and knowledge work. Anthropic reports that one early tester finished a 680,000-line code migration with it in under a day.

Which reasoning levels does Claude Opus 5.5 support in Super Agent?

Low, medium, high, extra high and max. Thinking is always on, and Super Agent starts it at medium.

Claude Opus 5.5 or Claude Fable 5.1?

Anthropic's docs suggest starting with Opus 5.5 for most work and moving to Fable 5.1 when Opus 5.5 at higher effort still falls short. Both sit in the featured section of the model picker.

More chat models on Flashloop

Browse all models