Gangsta AI

The Proof: AI models, graded head-to-head

10 questions, one verifiable answer each. An impartial top model graded every answer 0–100. Consensus got all 10 right — matching the very best model without you having to know which one that is — while ChatGPT, the most-used default, got 8. Here is every result.

Run your own question free →

Consensus vs ChatGPT (the most-used default)

Ask Consensus: 10/10 correct — never scored below 98. ChatGPT: 8/10. No guessing which model to trust.

Correct answers, by model

Ask Consensus10/10
Grok10/10
Claude Sonnet9/10
Gemini Flash9/10
ChatGPT8/10
Mistral4/10

Any single model can ace one set and flunk the next — and you can't tell which in advance. Consensus runs them all and returns the best answer every time.

Every question, every answer

Round 1

Which is larger: 9.11 or 9.9?

Answer: 9.9 is larger. 9.9 = 9.90 and 9.90 > 9.11. (Reading 9.11 as bigger because '11 > 9' is the classic error.)

  • ChatGPT: 9.9 is larger than 9.11. (100)
  • Claude Sonnet: 9.9 is larger than 9.11. When comparing decimals, you can think of them as: - 9.11 = 9.11 - 9.9 = 9.90 Since 90 hundredths is greater than 11 hundredths, 9.9 is the larger number. (100)
  • Gemini Flash: Let's compare the two numbers: * **9.11** * **9.9** Both numbers have the same whole number part, which is 9. So, we need to look at the decimal part. When comparing decimals, it's often helpful to ensure they have the same number of decimal places. We can add a zero to 9.9 without changing its value: * **9.11** * **9.90** Now, we compare the digits after the decimal point, from left to right: * The first digit after the decimal in 9.11 is **1**. * The first digit after the decimal in 9.90 is ** (100)
  • Grok: **9.9 is larger than 9.11.** To compare them directly: - \(9.9 = 9.90\) - \(9.11 = 9.11\) Since \(0.90 > 0.11\), it follows that \(9.9 > 9.11\). (If these were software version numbers rather than plain decimals, the comparison could differ by convention, but as ordinary real numbers, 9.9 is larger.) (100)
  • Ask Consensus: **Bottom line:** 9.9 is larger than 9.11 — by 0.79. Pad it to 9.90 vs 9.11 and the answer is obvious. ### Second Shot The answer is **9.9 > 9.11**. Here's the clean way to see it: - Both numbers share the same whole part: **9**. - Line up the decimal places by writing 9.9 as **9.90** (adding a trailing zero changes nothing). - Compare the tenths digit first: **9 tenths vs 1 tenth**. Nine wins. Done. - If you prefer hundredths: **90 hundredths vs 11 hundredths**. Same result. - The gap is **9.90 − 9.11 = 0.79**, so this isn't a close call. 📌 Why this trips people up: the eye reads "11" as bigg (100)

Round 2

Sort these four numbers from smallest to largest: 0.5, 0.55, 0.505, 0.5005.

Answer: 0.5, 0.5005, 0.505, 0.55. (As 4 decimals: 0.5000 < 0.5005 < 0.5050 < 0.5500.)

  • ChatGPT: To sort the numbers 0.5, 0.55, 0.505, and 0.5005 from smallest to largest, we can compare them by considering their decimal values: - 0.5 is equivalent to 0.5000. - 0.5005 is already given. - 0.505 is already given. - 0.55 is equivalent to 0.5500. Comparing the values: 1. 0.5 (0.5000) 2. 0.5005 3. 0.505 4. 0.55 (0.5500) Therefore, the numbers sorted from smallest to largest are: 0.5, 0.5005, 0.505, 0.55. (100)
  • Claude Sonnet: Looking at these numbers, I'll compare them digit by digit after the decimal point: - 0.5 = 0.5000 - 0.55 = 0.5500 - 0.505 = 0.5050 - 0.5005 = 0.5005 From smallest to largest: **0.5, 0.5005, 0.505, 0.55** (100)
  • Gemini Flash: Let's sort these four numbers from smallest to largest: 0.5, 0.55, 0.505, 0.5005 To compare these decimals, it's helpful to ensure they all have the same number of decimal places. We can do this by adding trailing zeros without changing their value. The number with the most decimal places is 0.5005 (four decimal places). So, let's rewrite them: 0.5000 0.5500 0.5050 0.5005 Now, let's compare them digit by digit from left to right: 1. **Compare the whole number part:** All are 0. 2. **Compare the (100)
  • Grok: 0.5, 0.5005, 0.505, 0.55 (100)
  • Mistral: To sort the numbers **0.5**, **0.55**, **0.505**, and **0.5005** from smallest to largest, follow these steps: 1. **Align the Decimal Places:** It's easier to compare the numbers when they have the same number of decimal places. Let's add zeros to each number to make them all four decimal places long: - 0.5 → **0.5000** - 0.55 → **0.5500** - 0.505 → **0.5050** - 0.5005 → **0.5005** 2. **Compare Each Digit from Left to Right:** - **First Decimal Place (Tenths):** All numbers have **5** in the ten (100)
  • Ask Consensus: **Bottom line:** 0.5 < 0.5005 < 0.505 < 0.55. Pad every number to four decimal places and the order falls out instantly. ### Second Shot The answer is **0.5, 0.5005, 0.505, 0.55**. The trap here is the eye: 0.5005 *looks* biggest because it has the most digits. Digit count means nothing. Place value is everything. The clean method — pad to equal length: - 0.5 → **0.5000** - 0.5005 → **0.5005** - 0.505 → **0.5050** - 0.55 → **0.5500** Now read them like whole numbers: 5000 < 5005 < 5050 < 5500. Done. Why that works, column by column: - **Tenths:** all four have a 5. No separation. - **Hundredth (100)

Round 3

Compute the exact product: 3127 × 419. Give the exact integer.

Answer: 1,310,213. (3127×400 = 1,250,800; 3127×19 = 59,413; sum = 1,310,213.)

  • ChatGPT: To compute the exact product of \( 3127 \times 419 \), we can use standard long multiplication: \[ \begin{array}{c} \phantom{+} 3127 \\ \times \phantom{+} 419 \\ \hline \end{array} \] First, multiply \( 3127 \) by \( 9 \): \[ \begin{array}{c} \phantom{+} 28143 \quad \text{(this is \( 3127 \times 9 \))} \\ \end{array} \] Next, multiply \( 3127 \) by \( 1 \) (which is actually \( 10 \) because it is in the tens place): \[ \begin{array}{c} \phantom{0} 3127\phantom{0} \quad \text{(this is \( 3127 \t (0)
  • Claude Sonnet: I need to compute 3127 × 419. Let me break this down: 3127 × 419 = 3127 × (400 + 19) = 3127 × 400 + 3127 × 19 First: 3127 × 400 = 1,250,800 Second: 3127 × 19 = 3127 × 20 - 3127 × 1 = 62,540 - 3,127 = 59,413 Total: 1,250,800 + 59,413 = 1,310,213 The exact product is **1,310,213**. (100)
  • Gemini Flash: To compute the exact product of 3127 and 419, we will perform multiplication. We can set up the multiplication as follows: 3127 x 419 ------ Step 1: Multiply 3127 by the units digit of 419, which is 9. 3127 x 9 ------ 28143 (9 * 3127) Step 2: Multiply 3127 by the tens digit of 419, which is 1 (representing 10). 3127 x 10 ------ 31270 (10 * 3127, or place a 0 and multiply by 1) Step 3: Multiply 3127 by the hundreds digit of 419, which is 4 (representing 400). 3127 x 400 ------ 1250800 (400 * 3127 (97)
  • Grok: The product can be computed by decomposition as follows: \[ 419 = 400 + 10 + 9, \] so \[ 3127 \times 419 = 3127 \times 400 + 3127 \times 10 + 3127 \times 9. \] First, \[ 3127 \times 400 = 1{,}250{,}800. \] Next, \[ 3127 \times 10 = 31{,}270. \] Then \[ 3127 \times 9 = 28{,}143. \] Adding these partial products yields \[ 1{,}250{,}800 + 31{,}270 = 1{,}282{,}070, \] \[ 1{,}282{,}070 + 28{,}143 = 1{,}310{,}213. \] The same result is obtained by the alternative decompositions \(419 = 400 + 19\) and (100)
  • Ask Consensus: **Bottom line:** 3127 × 419 = **1,310,213**. Three of four models land here; the lone outlier is an addition slip, not a real dispute. ### Second Shot The answer is **1,310,213**. The cleanest path is splitting 419 into 400 + 19: - **3127 × 400 = 1,250,800** - **3127 × 19** = 3127 × 20 − 3127 = 62,540 − 3,127 = **59,413** - **1,250,800 + 59,413 = 1,310,213** ✅ Same result by the textbook place-value route (400 + 10 + 9): - 3127 × 9 = 28,143 - 3127 × 10 = 31,270 - 3127 × 400 = 1,250,800 - 28,143 + 31,270 = 59,413; add 1,250,800 → **1,310,213** Two independent sanity checks confirm it: - **Digit (100)

Round 4

Is 3571 a prime number? Answer yes or no with a one-line justification.

Answer: Yes, 3571 is prime. It has no prime factor up to its square root (~59.8): not divisible by 2,3,5,7,11,...,59.

  • ChatGPT: Yes, 3571 is a prime number because it has no positive divisors other than 1 and itself. (90)
  • Claude Sonnet: No. 3571 = 59 × 61, so it is composite. (0)
  • Gemini Flash: Yes, 3571 is a prime number because it is only divisible by 1 and itself. (90)
  • Grok: Yes, 3571 is prime as it has no divisors other than 1 and itself among primes ≤ √3571 ≈ 59.7. (98)
  • Mistral: Yes, 3571 is a prime number because it has no positive divisors other than 1 and itself. (90)
  • Ask Consensus: **Bottom line:** Yes — 3571 is prime; no prime up to 59 (its square-root ceiling) divides it, and the only "factorization" offered, 59 × 61, actually equals 3599. ### Second Shot **Yes, 3571 is a prime number.** The one-line justification: **3571 has no prime factor ≤ √3571 ≈ 59.8**, so by trial division it is prime. The full check, for anyone who wants to see the work rather than take it on faith: - **2, 5:** odd, doesn't end in 0 or 5. - **3:** digit sum 3+5+7+1 = 16, not divisible by 3. - **7:** 7 × 510 = 3570, remainder 1. - **11:** 11 × 324 = 3564, remainder 7. - **13:** 13 × 274 = 3562, (100)

Round 5

Which is larger: 1/3 or 0.333?

Answer: 1/3 is larger. 1/3 = 0.333333... which exceeds 0.333.

  • ChatGPT: To compare \( \frac{1}{3} \) and 0.333, let's consider the decimal representation of \( \frac{1}{3} \). When \( \frac{1}{3} \) is converted into a decimal, it becomes 0.3333..., which is a repeating decimal where the digit 3 repeats indefinitely. On the other hand, 0.333 is a decimal that stops after three digits. Since \( \frac{1}{3} = 0.3333...\) and has repeating 3s, it is actually larger than 0.333 because 0.333 can be thought of as 0.3330, which is less than 0.3333.... Therefore, \( \frac{1 (96)
  • Claude Sonnet: 1/3 is larger. Here's why: 0.333 is an approximation, while 1/3 is the exact value. - 1/3 = 0.333333... (the 3s repeat infinitely) - 0.333 = 333/1000 The difference is small (about 0.000333...), but 1/3 is indeed larger than 0.333. (97)
  • Gemini Flash: To compare $1/3$ and $0.333$, we can convert $1/3$ to a decimal or convert $0.333$ to a fraction. Method 1: Convert $1/3$ to a decimal. To convert $1/3$ to a decimal, we perform the division: $1 \div 3 = 0.333333...$ (the 3s repeat infinitely). Now we compare $0.333333...$ with $0.333$. We can write $0.333$ as $0.333000...$. Comparing the digits from left to right: The first digit after the decimal point is 3 for both. The second digit after the decimal point is 3 for both. The third digit after (99)
  • Grok: **1/3 is larger than 0.333.** 1/3 = 0.333... (3 repeating forever), while 0.333 is exactly 333/1000. Thus \[ \frac{1}{3} - 0.333 = \frac{1}{3} - \frac{333}{1000} = \frac{1000 - 999}{3000} = \frac{1}{3000} > 0. \] So \( \frac{1}{3} > 0.333 \). (100)
  • Ask Consensus: **Bottom line:** 1/3 is larger. It equals 0.3333… repeating forever; 0.333 stops. The gap is exactly 1/3000. ### Second Shot **1/3 > 0.333.** Not by much, but it's not a tie, and it's not close to being one in a mathematical sense — the inequality is exact. Two clean ways to see it: - **Line up the decimals.** 1/3 = 0.333333… (the 3s never stop). Write 0.333 as 0.333000… Match digit by digit: the first three decimal places agree, then the fourth place is **3 vs 0**. First difference wins, so 1/3 is bigger. - **Put both over a common denominator.** 0.333 = 333/1000. 1/3 = 1000/3000, and 333/100 (100)

Round 6

How many times does the letter 'r' appear in the word 'refrigerator'?

Answer: 4 times. r-e-f-R-i-g-e-R-a-t-o-R plus the first R: refRigeratoR → R at positions 1, 4, 8, 12 = 4.

  • ChatGPT: The letter 'r' appears three times in the word 'refrigerator'. (0)
  • Claude Sonnet: To count the letter 'r' in the word 'refrigerator', I'll examine each letter: r-e-f-r-i-g-e-r-a-t-o-r The letter 'r' appears in positions: - 1st position: **r** - 4th position: **r** - 8th position: **r** - 12th position: **r** The letter 'r' appears **4 times** in the word 'refrigerator'. (100)
  • Gemini Flash: The letter 'r' appears **3** times in the word 'refrigerator'. (0)
  • Grok: The word "refrigerator" contains the letter 'r' 4 times. (100)
  • Mistral: Let's count the number of times the letter 'r' appears in the word **"refrigerator"**. Here is the word broken down with each 'r' highlighted: **r** e f **r** i g e **r** a t o **r** Counting them: 1. First 'r' 2. Second 'r' 3. Third 'r' 4. Fourth 'r' So, the letter 'r' appears **4 times** in "refrigerator". (100)
  • Ask Consensus: **Bottom line:** The letter **'r' appears 4 times** in "refrigerator" — positions 1, 4, 8, and 12. Any answer of three is a miscount. ### Second Shot The answer is **four**. Spell it out and the count is not up for debate: - **r**-e-f-**r**-i-g-e-**r**-a-t-o-**r** - 1st letter: **r** - 4th letter: **r** (after "ref") - 8th letter: **r** (after "refrige") - 12th letter: **r** (the final letter) That's four hits across a 12-letter word. The word is easy to undercount because two of the r's are in the middle of the word and get swallowed by the "frige" chunk when read at speed, and the pronunciat (100)

Round 7

What is the next number in this sequence, and state the rule in one line: 2, 3, 5, 9, 17, ...?

Answer: 33. Differences double (1, 2, 4, 8, then 16): 17 + 16 = 33. Equivalently a(n+1) = 2*a(n) − 1.

  • ChatGPT: The next number in the sequence is 33. The rule is: each term is the previous term plus the next power of 2 (e.g., 1, 2, 4, 8, 16, ...). (100)
  • Claude Sonnet: Based on the search results, **the next number is 33**. **Rule:** Each term equals the previous term plus the next power of 2 (differences are 1, 2, 4, 8, 16, ...). Source: [Filo - Complete the series: 2, 3, 5, 9, 17, ?](https://askfilo.com/user-question-answers-smart-solutions/complete-the-series-2-3-5-9-17-3431313339343233) (98)
  • Gemini Flash: The next number in the sequence is 33. The rule is that each term is the previous term plus the next power of 2, starting with 2^0 (1) added to the first term (2) to get the second term (3) (Filo, brainly.com). (96)
  • Grok: **33 (each term = previous + next power of 2)** The differences are +1, +2, +4, +8, … (i.e. successive powers of 2), so 17 + 16 = 33. (Equivalently, each term = 2 × previous − 1.) (100)
  • Ask Consensus: **Bottom line:** The next number is **33**. Rule: each term is **double the previous term minus 1** (equivalently, add successive powers of 2: +1, +2, +4, +8, +16). ### Second Shot The answer is **33**. Look at the gaps between terms: - 2 → 3: **+1** - 3 → 5: **+2** - 5 → 9: **+4** - 9 → 17: **+8** - 17 → ?: **+16** → **33** The differences are the powers of 2 in order (2⁰, 2¹, 2², 2³, 2⁴), so the next jump is 16 and 17 + 16 = 33. **Rule in one line:** aₙ = 2·aₙ₋₁ − 1 (each term is twice the previous term, minus one). Two things worth knowing about that rule: - ✅ It is **exactly equivalent** t (100)

Round 8

How many days are there from January 1, 2024 to March 1, 2024, counting both the start and end dates? Note 2024 is a leap year.

Answer: 61 days. Jan 31 + Feb 29 (leap year) + Mar 1 = 31 + 29 + 1 = 61, inclusive of both endpoints.

  • ChatGPT: To calculate the number of days from January 1, 2024, to March 1, 2024, including both start and end dates, you begin with January 1, 2024, and end with March 1, 2024. In a leap year, January has 31 days, and February has 29 days. - January 1 to January 31: 31 days - February 1 to February 29: 29 days - March 1: 1 day Adding these up, you get 31 + 29 + 1 = **61 days**. (100)
  • Claude Sonnet: Based on the search results provided, **from January 1, 2024 to March 1, 2024, counting both the start and end dates, there are 61 days**. Here's the calculation: - January has 31 days - February 2024 has 29 days (leap year) - March 1 is 1 day Total: 31 + 29 + 1 = **61 days** This is confirmed in the search results which state: "From January 1 to March 1, 2024, there are 61 days, including both start and end dates. 2024 is a leap year." (100)
  • Gemini Flash: There are 61 days from January 1, 2024 to March 1, 2024, counting both the start and end dates. 2024 is a leap year (Live web search results as of 2026-09-25). (97)
  • Grok: **61 days** 2024 is a leap year, so February has 29 days. Counting both endpoints: - January: 31 days (Jan 1–31) - February: 29 days - March 1: 1 day Total: \(31 + 29 + 1 = 61\). (March 1 is the 61st day of a leap year, so the inclusive span from day 1 to day 61 is 61 days.) (100)
  • Ask Consensus: **Bottom line:** **61 days.** January (31) + February in a leap year (29) + March 1 (1) = 61, counting both endpoints. ### Second Shot The answer is **61 days**. The count breaks down cleanly: - **January 1–31:** 31 days - **February 1–29:** 29 days (2024 is a leap year, so February gets the extra day) - **March 1:** 1 day **31 + 29 + 1 = 61.** A second way to sanity-check it: March 1 is the **61st day of the year** in a leap year (day-of-year 61). Counting inclusively from day 1 to day 61 gives exactly 61 days. Both methods land on the same number, so there's no ambiguity here. 📌 The "counti (100)

Round 9

A store sells apples at 3 for $1.35 and oranges at 5 for $2.75. If I buy 7 apples and 8 oranges, what is the exact total cost?

Answer: $7.55. Apples $0.45 each × 7 = $3.15; oranges $0.55 each × 8 = $4.40; total $7.55.

  • ChatGPT: To determine the total cost, we first need to calculate the cost of the apples and oranges separately. **Apples:** - The cost of apples is 3 for $1.35. - Therefore, the cost per apple is \( \frac{1.35}{3} = 0.45 \) dollars. - If you buy 7 apples, the total cost for the apples is \( 7 \times 0.45 = 3.15 \) dollars. **Oranges:** - The cost of oranges is 5 for $2.75. - Therefore, the cost per orange is \( \frac{2.75}{5} = 0.55 \) dollars. - If you buy 8 oranges, the total cost for the oranges is \( (100)
  • Claude Sonnet: I need to find the cost of 7 apples and 8 oranges. **Finding the cost per apple:** - 3 apples cost $1.35 - Cost per apple = $1.35 ÷ 3 = $0.45 **Finding the cost per orange:** - 5 oranges cost $2.75 - Cost per orange = $2.75 ÷ 5 = $0.55 **Calculating the total cost:** - 7 apples: 7 × $0.45 = $3.15 - 8 oranges: 8 × $0.55 = $4.40 - Total cost = $3.15 + $4.40 = **$7.55** (100)
  • Gemini Flash: Let's calculate the cost of apples and oranges separately. **Apples:** * Price of 3 apples = $1.35 * Price of 1 apple = $1.35 / 3 = $0.45 * Cost of 7 apples = 7 * $0.45 = $3.15 **Oranges:** * Price of 5 oranges = $2.75 * Price of 1 orange = $2.75 / 5 = $0.55 * Cost of 8 oranges = 8 * $0.55 = $4.40 **Total Cost:** * Total cost = Cost of apples + Cost of oranges * Total cost = $3.15 + $4.40 = $7.55 The exact total cost is **$7.55**. (100)
  • Grok: The apples are priced at 3 for $1.35, so the cost of a single apple is \(1.35 / 3 = 0.45\). Thus, the cost of 7 apples is \(7 \times 0.45 = 3.15\). The oranges are priced at 5 for $2.75, so the cost of a single orange is \(2.75 / 5 = 0.55\). Thus, the cost of 8 oranges is \(8 \times 0.55 = 4.40\). The combined total is \(3.15 + 4.40 = 7.55\). (The divisions are exact, as \(1.35 = 27/20\) and \(2.75 = 11/4\), so the per-fruit prices are the terminating decimals \(9/20\) and \(11/20\).) Final answ (100)
  • Ask Consensus: **Bottom line:** The exact total is **$7.55** — $3.15 for 7 apples plus $4.40 for 8 oranges. ### Second Shot The pricing here is clean, so the arithmetic is exact with no rounding involved. **Apples:** - 3 for $1.35 → **$0.45 each** (1.35 ÷ 3 = 0.45, exactly) - 7 × $0.45 = **$3.15** **Oranges:** - 5 for $2.75 → **$0.55 each** (2.75 ÷ 5 = 0.55, exactly) - 8 × $0.55 = **$4.40** **Total:** - $3.15 + $4.40 = **$7.55** ✅ Quick sanity check without decimals: in cents, apples are 135/3 = 45¢ and oranges are 275/5 = 55¢. 7 × 45 = 315¢, 8 × 55 = 440¢, sum 755¢. Same answer, no floating-point wobble. 📌 (98)

Round 10

For which specific achievement did Albert Einstein win his Nobel Prize in Physics, and in what year?

Answer: The photoelectric effect (the law of the photoelectric effect), awarded 1921. NOT for relativity — that is the common misconception.

  • ChatGPT: Albert Einstein was awarded the Nobel Prize in Physics in 1921. He received the prize for his discovery of the photoelectric effect, which was a pivotal contribution to the development of quantum theory. (92)
  • Claude Sonnet: Albert Einstein won the Nobel Prize in Physics in **1921** for his **discovery of the law of the photoelectric effect**. Interestingly, he did not win it for his theory of relativity, which is what he's most famous for today. The photoelectric effect work, published in 1905, demonstrated that light behaves as particles (photons) and was crucial in establishing quantum theory. The Nobel Committee considered this achievement more experimentally verified and less controversial at the time compared (100)
  • Gemini Flash: Albert Einstein won his Nobel Prize in Physics in **1921** for his **discovery of the law of the photoelectric effect**. (100)
  • Grok: **Albert Einstein won the 1921 Nobel Prize in Physics for his discovery of the law of the photoelectric effect.** The official Nobel citation reads: “for his services to Theoretical Physics, and especially for his discovery of the law of the photoelectric effect.” (He was awarded the 1921 prize in 1922; it was not given for relativity.) (100)
  • Mistral: Albert Einstein won the **Nobel Prize in Physics in 1921** (awarded in 1922) for his **explanation of the photoelectric effect**, a key contribution to quantum theory. His discovery demonstrated that light consists of discrete packets of energy (later called *photons*), which helped establish the particle-like nature of light and laid the foundation for quantum mechanics. Interestingly, the Nobel Committee did not award him the prize for his more famous theories of relativity (special or general (98)
  • Ask Consensus: **Bottom line:** The 1921 Nobel Prize in Physics, for his discovery of the **law of the photoelectric effect** — not relativity. The prize was actually handed over in 1922. ### Second Shot The answer is **1921**, and the specific achievement is the **law of the photoelectric effect**. The official citation reads: *"for his services to Theoretical Physics, and especially for his discovery of the law of the photoelectric effect."* Two details in that wording matter: - 📌 **"Law of" the photoelectric effect** — Einstein did not discover the effect itself. Hertz and Hallwachs observed it in 1887; (100)

How this was measured

Each question has one verifiable answer. Every model is asked via the live product; AskConsensus returns its verdict; an impartial top model grades each answer 0-100 against the verified ground truth.

Try Consensus on your hardest question →