DeepSeek R1 (Jan '25) vs MiniMax-M3
Which is better for the money: DeepSeek R1 (Jan '25) by DeepSeek or MiniMax-M3 by MiniMax? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-25.
Trade-off
Verdict
Depends on your budget
MiniMax-M3 scores 17.8 points higher (29.2 vs 11.4) but costs 1.2× as much to run the benchmark ($538 vs $461; premium vs mid-range cost classes). Pick MiniMax-M3 when the extra intelligence is worth the premium, otherwise DeepSeek R1 (Jan '25).
| Metric | DeepSeek R1 (Jan '25) | MiniMax-M3 | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 11.4 | 29.2 | +156.1% MiniMax-M3 is 2.6× more intelligent |
| Total AA cost (lower is better) | $460.74 | $537.98 | +16.8% MiniMax-M3 is 16.8% more expensive |
| End-to-end time (lower is better) | 20.16s | 25.25s | +25.2% MiniMax-M3 is 25.2% slower |
| Price per 1M tokens (in / out) | $2.00 / $4.00 | $0.30 / $1.20 | |
| Input modalities | Text | Text, Image, Video |
MiniMax-M3 vs DeepSeek R1 (Jan '25): 156.1% more intelligence, 16.8% more cost to run the AA benchmark, and 25.2% slower end-to-end.
Today's Pareto frontier
Against the frontier
- DeepSeek R1 (Jan '25) is off the frontier: at the same or lower cost, GPT-6.1 Sol (Medium) scores 47.80 for $361.
- MiniMax-M3 is off the frontier: at the same or lower cost, GPT-6.1 Sol (High) scores 50.20 for $521.
DeepSeek R1 (Jan '25)MiniMax-M3Today'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/deepseek-r1-0120/minimax-m3 · Markdown: /api/v1/compare/deepseek-r1-0120/minimax-m3?format=md · llms.txt · agent skill