MiniMax-M3 vs Step 5 Preview
Which is better for the money: MiniMax-M3 by MiniMax or Step 5 Preview by StepFun? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-24.
Trade-off
Verdict
Depends on your budget
Step 5 Preview scores 14.5 points higher (43.7 vs 29.2) but costs 2.2× as much to run the benchmark ($1190 vs $538; both premium class). Pick Step 5 Preview when the extra intelligence is worth the premium, otherwise MiniMax-M3.
| Metric | MiniMax-M3 | Step 5 Preview | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 29.2 | 43.7 | +49.7% Step 5 Preview is 49.7% more intelligent |
| Total AA cost (lower is better) | $537.98 | $1,190.26 | +121.2% Step 5 Preview is 2.2× more expensive |
| End-to-end time (lower is better) | 26.74s | 12.06s | −54.9% Step 5 Preview is 54.9% faster |
| Price per 1M tokens (in / out) | $0.30 / $1.20 | $1.00 / $2.70 | |
| Input modalities | Text, Image, Video | Text, Image |
Step 5 Preview vs MiniMax-M3: 49.7% more intelligence, 121.2% more cost to run the AA benchmark, and 54.9% faster end-to-end.
Today's Pareto frontier
Against the frontier
- MiniMax-M3 is off the frontier: at the same or lower cost, GPT-6.1 Sol (High) scores 50.20 for $521.
- Step 5 Preview is off the frontier: at the same or lower cost, GPT-6.1 Sol (Max) scores 51.80 for $1082.
MiniMax-M3Step 5 PreviewToday's Pareto frontierOpen-weights frontierEvery other model
Matchup
Compare something else
Agents:
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/minimax-m3/step-5 · Markdown: /api/v1/compare/minimax-m3/step-5?format=md · llms.txt · agent skill