Chat model

Gemini 4 Argon

low reasoning · 16 questions per request

96.4% of the 2,198 questions right (95.5% to 97.1%, 95% interval). Back to the leaderboard.

How we called it

Kind: Chat model
Access: Gemini API
Paid as: Provider service
Model ID: gemini-4-argon
Confidence from: Written estimate
Run: 557, 3 November 2026

Settings

Reasoning: low
Request format: 16 questions per request
Requests at once: 1
Temperature: not recorded
Harness seed: 0
Harness commit: ecf1af0326be
Prompt version: 254597b8be16

Thinking levels (Google's thinking documentation, read 28 October 2026): low, medium (default), high. Top level: high. This run: low.

Results

Questions	Accuracy	95% interval	Right
All questions	96.4%	95.5% to 97.1%	2,118 of 2,198
Reading	97.0%	95.3% to 98.0%	636 of 656
Grammar	96.1%	95.0% to 97.0%	1,482 of 1,542

Comparison group: cde. Compare with another setup.

Accuracy by topic

Topic	Part	Accuracy	Right
Word and phrase meanings	Reading	96.9%	127 of 131
Sentence meanings	Reading	100.0%	154 of 154
How a writer explains and develops ideas	Reading	90.0%	63 of 70
Topic and main point of a paragraph	Reading	99.0%	97 of 98
How a paragraph fits together	Reading	96.2%	102 of 106
Details that support a paragraph’s main point	Reading	95.9%	93 of 97
Word types, such as nouns and verbs	Grammar	96.4%	135 of 140
Groups of words built around a noun	Grammar	98.5%	64 of 65
Verbs and their forms	Grammar	100.0%	91 of 91
Verb forms used like nouns or modifiers	Grammar	98.6%	71 of 72
Word endings and what they do	Grammar	94.9%	75 of 79
How words are built	Grammar	95.7%	66 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	98.6%	73 of 74
Types of sentences	Grammar	96.5%	82 of 85
Questions combining several grammar topics	Grammar	91.8%	202 of 220
Sound patterns and changes	Grammar	94.8%	73 of 77
Spelling	Grammar	98.5%	131 of 133
Punctuation	Grammar	95.7%	112 of 117
Problems in sentence wording	Grammar	96.0%	168 of 175

Confidence

Answers kept at 99% or more: 1,842, of which 23 wrong (share right: 98.1% to 99.2%, 95% interval); 356 skipped, the model less than 99% sure
Calibration gap: 2.2 points
AUROC, all questions: 0.86 0.81 to 0.89
AUROC, reading: 0.96 0.94 to 0.98
AUROC, grammar: 0.84 0.80 to 0.88
Brier score: 0.032
Accuracy of its surest answers: most confident fifth: 99.9% right; most confident half: 99.3% right; most confident four fifths: 98.8% right
Typical score of a right answer: 99.0%
Typical score of a wrong answer: 95.0%

Repeats

We did not repeat this setup.

Latency

Latency, median request: 5.46 s (this run)
Latency, 90th percentile: 7.91 s
Requests at once: 1
Throughput: 152.3 questions per minute of request time
Request time: 14.4 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: $1.79
Whole run: $3.93
Basis: Estimated from published prices and recorded token usage
Price: $4 input, $0.20 cached input, $20 output per 1M tokens (source, 2026-10-28)
Input tokens per question: 349
Output tokens per question: 19