Command R7B (12-2024) Behavioral Fingerprint

How Cohere's Command R7B (12-2024) answers one-word probe questions at temperature 1.0 — random numbers, colors, animals, coin flips — measured over 364 valid samples in the open dataset of "One Token Is Enough" (arXiv:2607.10252).

Vendor

Cohere

Data generated

2026-07-08

Probe cells

14

Valid samples

364

Randomness score

67 / 100

Mean normalized entropy across cells, 0–100. 100 would be a uniform-random baseline; real models score far lower.

Signature answers

Favorite random number (1–100)

73

4 / 30 answers

Favorite color

red

4 / 28 answers

Coin flip: heads

61%

Share of heads across coin-flip cells.

Answer distributions by probe cell

Empirical distributions of normalized one-word answers, per task and language. H is the Shannon entropy in bits; the uniform baseline is log2 of the answer-domain size.

Coin flip · English

n = 15 · H = 0.97 bit · uniform baseline 1.00 bit

AnswerCountShare
tails960.0%
heads640.0%

Coin flip · Chinese

n = 16 · H = 0.70 bit · uniform baseline 1.00 bit

AnswerCountShare
heads1381.3%
tails318.8%

Favorite number · Chinese

n = 17 · H = 1.94 bit · uniform baseline 13.29 bit

AnswerCountShare
7952.9%
2423.5%
115.9%
315.9%
515.9%
1015.9%

Random animal · English

n = 27 · H = 4.13 bit · uniform baseline 5.64 bit

AnswerCountShare
a414.8%
elephant311.1%
owl27.4%
ostrich27.4%
tiger13.7%
jeffrey13.7%
penguin13.7%
otter13.7%

+ 12 more answers

Random animal · Chinese

n = 29 · H = 3.89 bit · uniform baseline 5.64 bit

AnswerCountShare
827.6%
大象26.9%
海豚26.9%
豚鼠13.4%
老虎13.4%
13.4%
河马13.4%
九龍象鱼13.4%

+ 12 more answers

Random city · English

n = 27 · H = 4.18 bit · uniform baseline 5.64 bit

AnswerCountShare
new311.1%
berlin311.1%
paris27.4%
toronto27.4%
melbourne27.4%
singapore13.7%
tokyo13.7%
santiago13.7%

+ 12 more answers

Random city · Chinese

n = 26 · H = 4.07 bit · uniform baseline 5.64 bit

AnswerCountShare
巴黎519.2%
大阪311.5%
华盛顿13.8%
edinburgh13.8%
ulturas13.8%
北京13.8%
伦敦13.8%
慕尼黑13.8%

+ 12 more answers

Random color · English

n = 28 · H = 3.66 bit · uniform baseline 4.91 bit

AnswerCountShare
red414.3%
orange414.3%
yellow414.3%
purple27.1%
turquoise27.1%
blue27.1%
brown27.1%
amber13.6%

+ 7 more answers

Random color · Chinese

n = 30 · H = 2.94 bit · uniform baseline 4.91 bit

AnswerCountShare
1240.0%
413.3%
绿26.7%
26.7%
粉红26.7%
26.7%
cyan13.3%
紫红13.3%

+ 4 more answers

Random letter · English

n = 30 · H = 3.67 bit · uniform baseline 4.70 bit

AnswerCountShare
q516.7%
n413.3%
a310.0%
b310.0%
m26.7%
p26.7%
s26.7%
l26.7%

+ 7 more answers

Random number 1-10 · English

n = 29 · H = 2.36 bit · uniform baseline 3.32 bit

AnswerCountShare
71034.5%
4517.2%
5517.2%
1413.8%
3413.8%
613.4%

Random number 1-10 · Chinese

n = 30 · H = 2.56 bit · uniform baseline 3.32 bit

AnswerCountShare
51033.3%
4620.0%
3516.7%
8413.3%
726.7%
213.3%
613.3%
913.3%

Random number 1-100 · English

n = 30 · H = 3.86 bit · uniform baseline 6.64 bit

AnswerCountShare
73413.3%
1310.0%
56310.0%
3726.7%
4226.7%
5326.7%
5426.7%
7626.7%

+ 8 more answers

Random number 1-100 · Chinese

n = 30 · H = 2.97 bit · uniform baseline 6.64 bit

AnswerCountShare
37930.0%
42620.0%
47413.3%
2326.7%
6726.7%
7326.7%
2513.3%
4513.3%

+ 3 more answers

Want to verify your API really serves Command R7B (12-2024)?

Point the free fingerprint checker at your endpoint: it samples the same probe questions from your browser and compares the distributions against this reference. Your API key never leaves your browser.

Verify your endpoint

Data source & license

Distributions: Tomáš Bruckner, "One Token Is Enough" (arXiv:2607.10252), official dataset Zenodo DOI 10.5281/zenodo.21278557, licensed CC-BY-4.0. Collected via OpenRouter at temperature 1.0 under the paper's fixed minimal one-word system prompt; answer keys are re-normalized with the checker's pipeline (color aliases, number words, coin h/t) so live probes are directly comparable.

Paper on arXivDataset on ZenodoCC-BY-4.0 license

More Cohere model fingerprints

Browse all 167 model fingerprints

Command R7B (12-2024) Behavioral Fingerprint — Favorite Random Numbers & Distribution | Tosea.ai