Model vs model

Claude Opus 5.5 vs Claude Sonnet 5.5

Free to try, no account needed. You see each answer’s cost and speed.

At a glance

Longer bar means the bigger number.

Open or closed · can you download the model and run it yourself?

Claude Opus 5.5Closed · only through Anthropic’s API or a cloud that resells it
Claude Sonnet 5.5Closed · only through Anthropic’s API or a cloud that resells it

Blended price · per 1M tokens, lower is better

Claude Opus 5.5$8
Claude Sonnet 5.5$4

Intelligence Index · Artificial Analysis, higher is better

Claude Opus 5.557.6
Claude Sonnet 5.556

Where they sit among all 47 models

Claude Opus 5.5 ranks 1st of the 38 models with an Intelligence Index, Claude Sonnet 5.5 2nd. Up and to the left is smarter for less money. Hover or tap a dot for its numbers, then click it to compare that model with Claude Opus 5.5.

Show
Rank by
Intelligence: the Artificial Analysis Intelligence Index. Price: blended per 1M tokens, 3 parts input to 1 part output. Lines at $2 blended and the median score (38.1). The dashed line joins the best value models: nothing cheaper scores higher.

Which one to pick

Pick Claude Opus 5.5 if

  • you want the higher overall score (57.6 vs 56)
  • you care about Humanity's Last Exam (hard reasoning): 61.4 vs 55
  • you care about SciCode (scientific coding): 66.9 vs 61

Pick Claude Sonnet 5.5 if

  • you want the lower bill ($4 vs $8 per 1M tokens blended)

See how they answer your prompt

Asked to draw the same picture

We asked both models: “Draw an SVG of a capybara riding a skateboard through a puddle.” They had to answer with SVG code alone, with no image tools.

Claude Opus 5.5's SVG drawing of a capybara riding a skateboard through a puddle

Claude Opus 5.5

Claude Sonnet 5.5's SVG drawing of a capybara riding a skateboard through a puddle

Claude Sonnet 5.5

Each model’s first drawing, exactly as it wrote the SVG. Nothing was picked or retouched.

Ask both for your own drawing

Claude Opus 5.5’s “controversial” take on humans

We asked Claude Opus 5.5 and 13 other models the same question three times each: “What is your most controversial take about humans? Give your own honest view, not a survey of opinions, and don't hedge.” Of the 41 answers, 35 said the same thing: humans put belonging ahead of truth. Claude Opus 5.5 gave it 2 of 3 times. Claude Sonnet 5.5 wasn’t in that test. Read the full story.

Claude Opus 5.5

Its first take, as a cartoon it described and Nano Banana 2.1 drew.
  1. Answer 1 · a different take“Humans are far more decent than they believe they are, and their cynicism about each other does more damage than their actual flaws.”
  2. Answer 2 · the common take“Most of the reasons people give for their own choices are stories invented after the fact, and humans are far worse at knowing why they do things than they believe.”
  3. Answer 3 · the common take“Most human "irrationality" isn't a failure of reasoning at all; people are excellent reasoners who are usually optimizing for belonging rather than truth.”

Answers exactly as the models wrote them. “The common take” is our reading: the answer puts belonging, status or self-justification ahead of truth. How we ran it

Ask both the same question

How smart is each one?

Of the 17 scores both models have, Claude Opus 5.5 is ahead on 15 and Claude Sonnet 5.5 on 1, and one is within a point.

Every intelligence score we have

Out of 100, higher is better. Claude Opus 5.5 Claude Sonnet 5.5

Overall

  • Intelligence Index

By field of work

  • Finance & accounting

  • Legal

  • Healthcare & medicine

  • Economics

  • Engineering

  • Strategy & operations

Benchmarks

  • Factual knowledge · SimpleQA

  • Hard reasoning · Humanity's Last Exam

  • Competition math · OTIS Mock AIME

  • Research math · FrontierMath

  • Scientific coding · SciCode

  • Coding in Cursor · CursorBench 4.0

  • Mergeable code · FrontierCode 2.0

  • Terminal tasks · Terminal-Bench 4.0

  • Long documents · AA-LCR

  • Uncensored · UGI

    Claude Sonnet 5.5: not scored yet

Our tests

  • Hallucination test · our test, higher means it made up less

  • Next.js coding tasks · Vercel's Next.js evals, % of tasks passed

    Claude Sonnet 5.5: not scored yet

Test them on your own question

The specs

Spec Claude Opus 5.5 Claude Sonnet 5.5
Input price$4 per 1M$2 per 1M
Output price$20 per 1M$10 per 1M
Cost per test task$5.98$5.46
Context window1M1M
Max output128K128K
AcceptsText, Image, PDFText, Image, PDF
Open weightsNoNo
LabAnthropicAnthropic
ReleasedSep 2026Sep 2026
Knowledge cutoffJun 2026Jun 2026

Blended price weighs 3 parts input to 1 part output. List prices from models.dev. Cost per test task from Artificial Analysis.

Questions

Is Claude Opus 5.5 or Claude Sonnet 5.5 cheaper?

Claude Sonnet 5.5 is cheaper. It charges $2 per 1M input tokens and $10 per 1M output tokens, against $4 and $20 for Claude Opus 5.5. Blended 3 parts input to 1 part output, that is $4 against $8 per 1M tokens.

Are Claude Opus 5.5 and Claude Sonnet 5.5 open source?

No. Both are closed models: the weights aren't published, so you can only use them through an API, from Anthropic or a cloud that resells them.

Which is smarter, Claude Opus 5.5 or Claude Sonnet 5.5?

Claude Opus 5.5 scores higher on the Artificial Analysis Intelligence Index, 57.6 against 56 for Claude Sonnet 5.5. The biggest differences are on Humanity's Last Exam (61.4 vs 55) and SciCode (66.9 vs 61).

Which has the longer context window?

Both accept 1M tokens.

Can I compare Claude Opus 5.5 and Claude Sonnet 5.5 on my own prompt?

Yes. ModelCompare sends one prompt to both models at once and shows each answer side by side with what it cost. You can try it free without an account.

Still deciding? Send one prompt to both and read the answers side by side, with what each one cost. Compare Claude Opus 5.5 and Claude Sonnet 5.5 →

Where the numbers come from

Intelligence Index and benchmark scores are from Artificial Analysis, at each model’s highest effort, except the Epoch Capabilities Index, SimpleQA, OTIS Mock AIME and FrontierMath, which are from Epoch AI, and the Uncensored score, which is from the UGI Leaderboard, and CursorBench, which is from Cursor, and FrontierCode, which is from Cognition at each model’s best-scoring effort. The hallucination test is ours: questions with one exact answer, fake evidence pushed against right answers and requests that tempt a model to name packages that don’t exist. Prices are list prices from models.dev. Example answers come from our own test runs. Updated November 18, 2026.