Model comparison
DeepSeek V4 Pro (Max) vs MiniMax M3
Head-to-head evidence from 23 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Verified leaderboard positions: DeepSeek V4 Pro (Max) #9; MiniMax M3 #18
BenchAlign evidence: DeepSeek V4 Pro (Max) not scored; MiniMax M3 supported. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. DeepSeek V4 Pro (Max) and MiniMax M3 share 23 comparable benchmark results. 3 of 8 categories are comparable. 19 results are unique to DeepSeek V4 Pro (Max); 22 to MiniMax M3.
Updated July 16, 2026- Shared results
- 23
- DeepSeek V4 Pro (Max) only
- 19
- MiniMax M3 only
- 22
- Comparable categories
- 3 / 8
Pick DeepSeek V4 Pro (Max) if you want the stronger benchmark profile. MiniMax M3 only becomes the better choice if coding is the priority or you would rather avoid the extra latency and token burn of a reasoning model.
Confidence note. This is a partial-evidence comparison with 23 shared benchmark results across 6 evidence categories; 3 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
DeepSeek V4 Pro (Max) is clearly ahead on the provisional aggregate, 76 to 70. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
DeepSeek V4 Pro (Max)'s sharpest advantage is in mathematics, where it averages 95.2 against 85.7. The single biggest benchmark swing on the page is SWE-bench Pro, 55.4% to 59%. MiniMax M3 does hit back in coding, so the answer changes if that is the part of the workload you care about most.
MiniMax M3 is also the more expensive model on tokens at $0.30 input / $1.20 output per 1M tokens, versus $0.43 input / $0.87 output per 1M tokens for DeepSeek V4 Pro (Max). DeepSeek V4 Pro (Max) is the reasoning model in the pair, while MiniMax M3 is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use.
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
| Category | DeepSeek V4 Pro (Max) | Δ | MiniMax M3 |
|---|---|---|---|
| Math | DeepSeek V4 Pro (Max)95.2 | Margin← 9.5 | MiniMax M385.7 |
| Agentic | DeepSeek V4 Pro (Max)74.5 | Margin← 2.2 | MiniMax M372.3 |
| Coding | DeepSeek V4 Pro (Max)70.9 | Margin→ 1.3 | MiniMax M372.2 |
| Knowledge | DeepSeek V4 Pro (Max)60.4 | MarginNo overlap | MiniMax M3Not measured |
| Multimodal | DeepSeek V4 Pro (Max)Not measured | MarginNo overlap | MiniMax M364.9 |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
SWE-bench Pro
CodingA 55.4%B 59%Winner: MiniMax M3Δ 3.6SWE-bench Pro: DeepSeek V4 Pro (Max) scored 55.4%; MiniMax M3 scored 59%. MiniMax M3 wins this benchmark. - Source ↗
Terminal-Bench 2.0
AgenticA 67.9%B 66%Winner: DeepSeek V4 Pro (Max)Δ 1.9Terminal-Bench 2.0: DeepSeek V4 Pro (Max) scored 67.9%; MiniMax M3 scored 66%. DeepSeek V4 Pro (Max) wins this benchmark. - Source ↗
BrowseComp
AgenticA 83.4%B 83.5%Winner: MiniMax M3Δ 0.1BrowseComp: DeepSeek V4 Pro (Max) scored 83.4%; MiniMax M3 scored 83.5%. MiniMax M3 wins this benchmark. - Source ↗
SWE-bench Verified
CodingA 80.6%B 80.5%Winner: DeepSeek V4 Pro (Max)Δ 0.1SWE-bench Verified: DeepSeek V4 Pro (Max) scored 80.6%; MiniMax M3 scored 80.5%. DeepSeek V4 Pro (Max) wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | DeepSeek V4 Pro (Max) | MiniMax M3 | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | DeepSeek V4 Pro (Max)$0.435 input / $0.87 output | MiniMax M3$0.3 input / $1.2 output | DeepSeek V4 Pro (Max) has the lower combined listed price. |
| Generation speedtokens per second | DeepSeek V4 Pro (Max)Not available | MiniMax M3Not available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | DeepSeek V4 Pro (Max)Not available | MiniMax M3Not available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | DeepSeek V4 Pro (Max)1M | MiniMax M31M | Listed context windows are equal. |
Benchmark Deep Dive
AgenticDeepSeek V4 Pro (Max) wins19 benchmarks
| Benchmark | DeepSeek V4 Pro (Max) | MiniMax M3 | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | 67.9% | 66% | DeepSeek V4 Pro (Max) leads |
| BrowseCompSource | 83.4% | 83.5% | MiniMax M3 leads |
| HLE w/ toolsSource | 48.2% | — | Not comparable |
| MCP AtlasSource | 73.6% | 74.2% | MiniMax M3 leads |
| GDPval-AASource | 1307 | 1395 | MiniMax M3 leads |
| ToolathlonSource | 51.8% | — | Not comparable |
| AA Agentic IndexSource | 36.4% | 35.4% | DeepSeek V4 Pro (Max) leads |
| APEX-Agents-AASource | 24.3% | — | Not comparable |
| τ²-bench resultsSource | 96.2% | 88.9% | DeepSeek V4 Pro (Max) leads |
| GDPval-AASource | 40.4% | 44.7% | MiniMax M3 leads |
| OSWorld-VerifiedSource | — | 70.1% | Not comparable |
| Claw-EvalSource | — | 74.5% | Not comparable |
| GDPval rubricsSource | — | 74.7% | Not comparable |
| BankerToolBenchSource | — | 76.1% | Not comparable |
| ResearchClawBenchSource | — | 19.8% | Not comparable |
| OSWorld 2.0Source | — | 4.6% | Not comparable |
| AA BriefcaseSource | — | 1110 | Not comparable |
| AA EnterpriseOps-GymSource | — | 32.1% | Not comparable |
| AA Harvey LABSource | — | 6.7% | Not comparable |
CodingMiniMax M3 wins15 benchmarks
| Benchmark | DeepSeek V4 Pro (Max) | MiniMax M3 | Result |
|---|---|---|---|
| LiveCodeBench Pass@1-COTSource | 93.5% | — | Not comparable |
| CodeforcesSource | 3206.0 | — | Not comparable |
| SWE-bench VerifiedSource | 80.6% | 80.5% | DeepSeek V4 Pro (Max) leads |
| SWE-bench ProSource | 55.4% | 59% | MiniMax M3 leads |
| SWE MultilingualSource | 76.2% | — | Not comparable |
| Terminal-Bench 2.0Source | 67.9% | 66.0% | DeepSeek V4 Pro (Max) leads |
| Vibe Code BenchSource | 49.93% | — | Not comparable |
| AA Coding IndexSource | 59.4% | 58.6% | DeepSeek V4 Pro (Max) leads |
| Terminal-Bench HardSource | 46.2% | 42.4% | DeepSeek V4 Pro (Max) leads |
| AA-SciCodeSource | 50.0% | 45.4% | DeepSeek V4 Pro (Max) leads |
| NL2RepoSource | — | 42.1% | Not comparable |
| VIBE V2Source | — | 50.1% | Not comparable |
| SVG-BenchSource | — | 63.7% | Not comparable |
| KernelBench HardSource | — | 28.8% | Not comparable |
| AA Terminal-Bench 2.1Source | — | 65.2% | Not comparable |
Reasoning4 benchmarks
Knowledge13 benchmarks
| Benchmark | DeepSeek V4 Pro (Max) | MiniMax M3 | Result |
|---|---|---|---|
| MMLU-ProSource | 87.5% | — | Not comparable |
| SimpleQASource | 57.9% | — | Not comparable |
| Chinese-SimpleQASource | 84.4% | — | Not comparable |
| GPQASource | 90.1% | — | Not comparable |
| GPQA-DSource | 90.1% | — | Not comparable |
| HLESource | 37.7% | — | Not comparable |
| Artificial Analysis Intelligence IndexSource | 44.3% | 44.4% | MiniMax M3 leads |
| AA-GPQA DiamondSource | 88.8% | 92.9% | MiniMax M3 leads |
| AA-HLESource | 35.9% | 37.1% | MiniMax M3 leads |
| AA-Omniscience IndexSource | -10.0% | 1.4% | MiniMax M3 leads |
| AA-Omniscience AccuracySource | 43.3% | 15.0% | DeepSeek V4 Pro (Max) leads |
| AA-Omniscience Hallucination RateSource | 94.0% | 16.1% | MiniMax M3 leads |
| AA Openness IndexSource | — | 33.3% | Not comparable |
MathDeepSeek V4 Pro (Max) wins5 benchmarks
Multimodal7 benchmarks
| Benchmark | DeepSeek V4 Pro (Max) | MiniMax M3 | Result |
|---|---|---|---|
| Design Arena WebsiteSource | 1269 | 1294 | MiniMax M3 leads |
| OfficeQA ProSource | — | 45.1% | Not comparable |
| OmniDocBench 1.5Source | — | 91.6% | Not comparable |
| MMMU-ProSource | — | 78.1% | Not comparable |
| VideoMMMUSource | — | 84.6% | Not comparable |
| Video-MME (with subtitle)Source | — | 85.4% | Not comparable |
| AA-MMMU-ProSource | — | 78.6% | Not comparable |
Inst. Following1 benchmarks
| Benchmark | DeepSeek V4 Pro (Max) | MiniMax M3 | Result |
|---|---|---|---|
| AA-IFBenchSource | 76.5% | 82.9% | MiniMax M3 leads |
Frequently Asked Questions (4)
Which is better, DeepSeek V4 Pro (Max) or MiniMax M3?
DeepSeek V4 Pro (Max) is ahead on BenchLM's provisional leaderboard, 76 to 70. The biggest single separator in this matchup is SWE-bench Pro, where the scores are 55.4% and 59%.
Which is better for coding, DeepSeek V4 Pro (Max) or MiniMax M3?
MiniMax M3 has the edge for coding in this comparison, averaging 72.2 versus 70.9. Inside this category, AA-SciCode is the benchmark that creates the most daylight between them.
Which is better for math, DeepSeek V4 Pro (Max) or MiniMax M3?
DeepSeek V4 Pro (Max) has the edge for math in this comparison, averaging 95.2 versus 85.7. MiniMax M3 stays close enough that the answer can still flip depending on your workload.
Which is better for agentic tasks, DeepSeek V4 Pro (Max) or MiniMax M3?
DeepSeek V4 Pro (Max) has the edge for agentic tasks in this comparison, averaging 74.5 versus 72.3. Inside this category, GDPval-AA is the benchmark that creates the most daylight between them.
Related Comparisons
Explore More
The AI models change fast. We track them for you.
A weekly brief for engineers and researchers covering new models, ranking shifts, and pricing changes.
Free. No spam. Unsubscribe anytime.