DeepSeek V3 (Dec '24) vs MiniMax-M3
Which is better for the money: DeepSeek V3 (Dec '24) by DeepSeek or MiniMax-M3 by MiniMax? 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
MiniMax-M3 scores 20.7 points higher (29.2 vs 8.5) but costs 21× as much to run the benchmark ($538 vs $25.4; premium vs budget cost classes). They are not alternatives at one budget: pick MiniMax-M3 when you need intelligence above 8.5, and DeepSeek V3 (Dec '24) when 8.5 is enough.
| Metric | DeepSeek V3 (Dec '24) | MiniMax-M3 | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 8.5 | 29.2 | +243.5% MiniMax-M3 is 3.4× more intelligent |
| Total AA cost (lower is better) | $25.37 | $537.98 | +2020.5% MiniMax-M3 is 21× more expensive |
| End-to-end time (lower is better) | 11.89s | 26.74s | +124.9% MiniMax-M3 is 124.9% slower |
| Price per 1M tokens (in / out) | $0.32 / $0.89 | $0.30 / $1.20 | |
| Input modalities | Text | Text, Image, Video |
MiniMax-M3 vs DeepSeek V3 (Dec '24): 243.5% more intelligence, 2020.5% more cost to run the AA benchmark, and 124.9% slower end-to-end.
Today's Pareto frontier
Against the frontier
- DeepSeek V3 (Dec '24) is off the frontier: at the same or lower cost, GPT-6 Luna (Low) scores 21.50 for $10.6.
- MiniMax-M3 is off the frontier: at the same or lower cost, GPT-6.1 Sol (High) scores 50.20 for $521.
DeepSeek V3 (Dec '24)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-v3/minimax-m3 · Markdown: /api/v1/compare/deepseek-v3/minimax-m3?format=md · llms.txt · agent skill