Chat model

Claude Opus 5.5

low reasoning · 16 questions per request

90.4% of the 2,198 questions right (89.1% to 91.6%, 95% interval). Back to the leaderboard.

We also ran this setup at 1 question per request: Claude Opus 5.5 · low reasoning · 1 question per request.

Which decision model can go first in front of it? See the Routing tab.

How we called it

Kind: Chat model
Access: Claude Code CLI, Claude subscription
Paid as: Subscription
Model ID: claude-opus-5-5
Confidence from: Written estimate
Run: 47, 4 October 2026

Settings

Reasoning: low
Request format: 16 questions per request
Requests at once: 1
Temperature: not recorded
Harness seed: 0
Harness commit: not recorded
Prompt version: not recorded

Effort levels (Anthropic's effort documentation, read 28 October 2026): low, medium (default), high, xhigh, max. Top level: max. This run: low.

Results

Questions	Accuracy	95% interval	Right
All questions	90.4%	89.1% to 91.6%	1,988 of 2,198
Reading	94.8%	92.8% to 96.3%	622 of 656
Grammar	88.6%	86.9% to 90.1%	1,366 of 1,542

Comparison group: jk. Compare with another setup.

Accuracy by topic

Topic	Part	Accuracy	Right
Word and phrase meanings	Reading	95.4%	125 of 131
Sentence meanings	Reading	96.8%	149 of 154
How a writer explains and develops ideas	Reading	85.7%	60 of 70
Topic and main point of a paragraph	Reading	98.0%	96 of 98
How a paragraph fits together	Reading	95.3%	101 of 106
Details that support a paragraph’s main point	Reading	93.8%	91 of 97
Word types, such as nouns and verbs	Grammar	97.9%	137 of 140
Groups of words built around a noun	Grammar	98.5%	64 of 65
Verbs and their forms	Grammar	94.5%	86 of 91
Verb forms used like nouns or modifiers	Grammar	97.2%	70 of 72
Word endings and what they do	Grammar	94.9%	75 of 79
How words are built	Grammar	84.1%	58 of 69
The roles of words and phrases in a sentence	Grammar	95.9%	139 of 145
How verbs show who acts and who receives the action	Grammar	95.9%	71 of 74
Types of sentences	Grammar	92.9%	79 of 85
Questions combining several grammar topics	Grammar	74.1%	163 of 220
Sound patterns and changes	Grammar	70.1%	54 of 77
Spelling	Grammar	83.5%	111 of 133
Punctuation	Grammar	84.6%	99 of 117
Problems in sentence wording	Grammar	91.4%	160 of 175

Confidence

Answers kept at 99% or more: 0, of which 0 wrong; 2,198 skipped, the model less than 99% sure
Calibration gap: 16.5 points
AUROC, all questions: 0.84 0.81 to 0.87
AUROC, reading: 0.88 0.84 to 0.92
AUROC, grammar: 0.82 0.79 to 0.86
Brier score: 0.101
Accuracy of its surest answers: most confident fifth: 99.8% right; most confident half: 98.0% right; most confident four fifths: 95.8% right
Typical score of a right answer: 80.0%
Typical score of a wrong answer: 50.0%

Repeats

Run 5 times on a sample of 200 questions.

Answers changed in 5 runs: 20 of 200
Same wrong answer in all 5: 2 of 200
Wrong in at least one run: 22 of 200
Accuracy in each run: 94.0%, 95.0%, 95.0%, 96.0%, 97.5%

Latency

Latency, median request: 13.4 s, measured on a sample (200 questions) over 5 runs
Latency, 90th percentile: 19.8 s
Requests at once: 1
Throughput: 71.9 questions per minute of request time
Request time: 30.7 minutes, over the requests whose answers were kept

Latency: the time from sending one request to receiving its answer, measured from our computer, network and any retries within the request included. Request time: the time at least one request was in flight, over the requests whose answers were kept. It is not the run's elapsed time: it leaves out requests that failed and any pause before a run was resumed.

Cost

$ per 1,000 questions: $4.93
Whole run: $10.90
Basis: Subscription run priced as if we had used the API
Price: $4 input, $0.20 cached input, $5 5-minute cache writes, $8 1-hour cache writes, $20 output per 1M tokens (source, 2026-10-08)
Input tokens per question: 565
Output tokens per question: 78