Benchmark menu

Model detail

Minimax M3

A static, shareable model page that remains complete without JavaScript.

Minimax M3

Provider: MiniMax · Public slug: minimax-minimax-m3

Rank
22
Rank movement
Baseline
Score
29.1
Last measured
2026-07-05
Coverage
XHigh not supported

Reasoning scores

XHigh
Medium
19.7
None
38.4

Generation health

Medium0%88%12%n 8None50%25%25%n 8 XHigh not supportedHigh repair rateHigh failure rate

Cost summary

Mean standard list-price equivalent
$0.05
Median unique-cell cost
$0.05
Unique-cell range
$0.0064–$0.10
Coverage
16 unique cells / 16 observed runs
Observed-run total
$0.79
Evidence basis
Provider-reported
Mean output
37k tok
Output range
3.5k–78k tok
Token samples
16

Cost means equal-weight unique model/game/reasoning cells. Provider-account billed spend is reported separately on Charts.

Run setup

  • Overall Models: combines public reasoning variants
  • Standard GameBench code-generation prompt