OverpayingForAIPricing desk

Reasoning · 1 checks · max 800 tokens · benched 2026-09-16

MiniMax M3 vs MiniMax M2.7 vs MiniMax M2.5 on count sellable units

Minimax models side by side on "Count sellable units": MiniMax M3 scores 10/10; MiniMax M3 is the cheapest answer scoring 8+ at $0.20 per 1,000 runs. Outputs, checks, judge reasons, latency and cost.

The prompt every model received

System

Think carefully, then give only the final answer in the requested format.

User

A warehouse receives 3 pallets. Each pallet holds 48 boxes and each box holds 12 units. On inspection, 15% of the boxes are damaged; round the number of damaged boxes down to a whole box. The damaged boxes are discarded. Then 7 undamaged boxes are sent to the quality lab and are not sold. How many units are left to sell? Reply with only the number.

Rubric for the judge: The single correct answer 1392 (144 boxes, 21 damaged, 7 to the lab, 116 x 12).

Side by side

Every cell is one OpenRouter call at temperature 0 with the prompt's token cap and reasoning effort "low" where the model supports it. Cost is usage × the catalogue rate in models.json. Quality is one judge call to anthropic/claude-haiku-4.5 against the prompt's rubric, cached per prompt version.

MiniMax M3

minimax/minimax-m3

10/10
top judge scorecheapest scoring 8+
Latency
384ms
Cost
$0.00020
Per 1,000
$0.20

267 in · 129 out (124 reasoning) · 1 words · checks 1/1

  • exact
1392

Judge: The output matches the correct answer of 1392 units exactly, provided in the required format of a bare number only.

MiniMax M2.7

minimax/minimax-m2.7

10/10
Latency
2.1s
Cost
$0.00075
Per 1,000
$0.75

111 in · 600 out (597 reasoning) · 1 words · checks 1/1

  • exact


1392

Judge: The output provides exactly the correct numerical answer (1392) in the bare format requested, matching the rubric perfectly.

MiniMax M2.5

minimax/minimax-m2.5

10/10
Latency
143ms
Cost
$0.00044
Per 1,000
$0.44

113 in · 377 out (355 reasoning) · 1 words · checks 1/1

  • exact
1392

Judge: The output matches the correct answer of 1392 units exactly, provided in the required format of a bare number only.

Frequently asked

What does this prompt test?

Reasoning: The single correct answer 1392 (144 boxes, 21 damaged, 7 to the lab, 116 x 12). The deterministic checks are exact.

Which model should I pick for this task?

If the judge's bar of 8/10 is good enough for you, MiniMax M3 at $0.20 per 1,000 runs. If you need the top score, MiniMax M3 at $0.20 per 1,000 runs.

If our calculators helped you cut down on hidden AI wallet leaks, thanks for using them. A tiny fraction of your savings is what keeps our pricing indexes updated daily.

Not sure which AI is cheapest for your use case? Find out in 30 seconds — no signup required.

AI cost intelligence

Stop overpaying for AI tools

Join the OverpayingForAI list for pricing updates, cheaper alternatives, and practical buying guidance.

Now tracking 50+ AI tools, models, platforms, subscriptions, coding tools, and automation products.

We use your email only for OverpayingForAI updates. Unsubscribe anytime.