Model comparison
MiniMax M3 vs Qwen 3.6 Max (preview)
Head-to-head evidence from 16 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Verified leaderboard positions: MiniMax M3 #18; Qwen 3.6 Max (preview) unranked
BenchAlign evidence: MiniMax M3 supported; Qwen 3.6 Max (preview) supported. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. MiniMax M3 and Qwen 3.6 Max (preview) share 16 comparable benchmark results. 3 of 8 categories are comparable. 29 results are unique to MiniMax M3; 6 to Qwen 3.6 Max (preview).
Updated July 16, 2026- Shared results
- 16
- MiniMax M3 only
- 29
- Qwen 3.6 Max (preview) only
- 6
- Comparable categories
- 3 / 8
Pick MiniMax M3 if you want the stronger benchmark profile. Qwen 3.6 Max (preview) only becomes the better choice if you want the stronger reasoning-first profile.
Confidence note. This is a partial-evidence comparison with 16 shared benchmark results across 5 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
MiniMax M3 has the cleaner provisional overall profile here, landing at 70 versus 68. It is a real lead, but still close enough that category-level strengths matter more than the headline number.
MiniMax M3's sharpest advantage is in mathematics, where it averages 85.7 against 18.4. The single biggest benchmark swing on the page is SWE-bench Pro, 59% to 57.3%.
Qwen 3.6 Max (preview) 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. MiniMax M3 gives you the larger context window at 1M, compared with 256K for Qwen 3.6 Max (preview).
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 | MiniMax M3 | Δ | Qwen 3.6 Max (preview) |
|---|---|---|---|
| Math | MiniMax M385.7 | Margin← 67.3 | Qwen 3.6 Max (preview)18.4 |
| Coding | MiniMax M372.2 | Margin← 21.2 | Qwen 3.6 Max (preview)51.0 |
| Agentic | MiniMax M372.3 | Margin← 6.9 | Qwen 3.6 Max (preview)65.4 |
| Knowledge | MiniMax M3Not measured | MarginNo overlap | Qwen 3.6 Max (preview)73.9 |
| Multimodal | MiniMax M364.9 | MarginNo overlap | Qwen 3.6 Max (preview)Not measured |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
SWE-bench Pro
CodingA 59%B 57.3%Winner: MiniMax M3Δ 1.7SWE-bench Pro: MiniMax M3 scored 59%; Qwen 3.6 Max (preview) scored 57.3%. MiniMax M3 wins this benchmark. - Source ↗
Terminal-Bench 2.0
AgenticA 66%B 65.4%Winner: MiniMax M3Δ 0.6Terminal-Bench 2.0: MiniMax M3 scored 66%; Qwen 3.6 Max (preview) scored 65.4%. MiniMax M3 wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | MiniMax M3 | Qwen 3.6 Max (preview) | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | MiniMax M3$0.3 input / $1.2 output | Qwen 3.6 Max (preview)Not available | A complete price comparison is not available. |
| Generation speedtokens per second | MiniMax M3Not available | Qwen 3.6 Max (preview)Not available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | MiniMax M3Not available | Qwen 3.6 Max (preview)Not available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | MiniMax M31M | Qwen 3.6 Max (preview)256K | MiniMax M3 lists the larger context window. |
Benchmark Deep Dive
AgenticMiniMax M3 wins18 benchmarks
| Benchmark | MiniMax M3 | Qwen 3.6 Max (preview) | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | 66% | 65.4% | MiniMax M3 leads |
| BrowseCompSource | 83.5% | — | Not comparable |
| OSWorld-VerifiedSource | 70.1% | — | Not comparable |
| MCP AtlasSource | 74.2% | — | Not comparable |
| Claw-EvalSource | 74.5% | — | Not comparable |
| AA Agentic IndexSource | 35.4% | — | Not comparable |
| τ²-bench resultsSource | 88.9% | 95.9% | Qwen 3.6 Max (preview) leads |
| GDPval-AASource | 44.7% | — | Not comparable |
| GDPval-AASource | 1395 | — | 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 |
| QwenClawBenchSource | — | 59.0% | Not comparable |
| QwenWebBenchSource | — | 1532 | Not comparable |
CodingMiniMax M3 wins12 benchmarks
| Benchmark | MiniMax M3 | Qwen 3.6 Max (preview) | Result |
|---|---|---|---|
| SWE-bench VerifiedSource | 80.5% | — | Not comparable |
| SWE-bench ProSource | 59% | 57.3% | MiniMax M3 leads |
| Terminal-Bench 2.0Source | 66.0% | 65.4% | MiniMax M3 leads |
| NL2RepoSource | 42.1% | 42.9% | Qwen 3.6 Max (preview) leads |
| AA Coding IndexSource | 58.6% | — | Not comparable |
| Terminal-Bench HardSource | 42.4% | 43.9% | Qwen 3.6 Max (preview) leads |
| AA-SciCodeSource | 45.4% | 46.9% | Qwen 3.6 Max (preview) leads |
| 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 |
| SciCodeSource | — | 47% | Not comparable |
Reasoning2 benchmarks
Knowledge8 benchmarks
| Benchmark | MiniMax M3 | Qwen 3.6 Max (preview) | Result |
|---|---|---|---|
| Artificial Analysis Intelligence IndexSource | 44.4% | 40.0% | MiniMax M3 leads |
| AA-GPQA DiamondSource | 92.9% | 88.8% | MiniMax M3 leads |
| AA-HLESource | 37.1% | 28.9% | MiniMax M3 leads |
| AA-Omniscience IndexSource | 1.4% | 10.2% | Qwen 3.6 Max (preview) leads |
| AA-Omniscience AccuracySource | 15.0% | 37.7% | Qwen 3.6 Max (preview) leads |
| AA-Omniscience Hallucination RateSource | 16.1% | 44.2% | MiniMax M3 leads |
| AA Openness IndexSource | 33.3% | — | Not comparable |
| SuperGPQASource | — | 73.9% | Not comparable |
MathMiniMax M3 wins3 benchmarks
Multimodal7 benchmarks
| Benchmark | MiniMax M3 | Qwen 3.6 Max (preview) | Result |
|---|---|---|---|
| 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 |
| Design Arena WebsiteSource | 1294 | — | Not comparable |
| AA-MMMU-ProSource | 78.6% | — | Not comparable |
Inst. Following1 benchmarks
| Benchmark | MiniMax M3 | Qwen 3.6 Max (preview) | Result |
|---|---|---|---|
| AA-IFBenchSource | 82.9% | 76.6% | MiniMax M3 leads |
Frequently Asked Questions (4)
Which is better, MiniMax M3 or Qwen 3.6 Max (preview)?
MiniMax M3 is ahead on BenchLM's provisional leaderboard, 70 to 68. The biggest single separator in this matchup is SWE-bench Pro, where the scores are 59% and 57.3%.
Which is better for coding, MiniMax M3 or Qwen 3.6 Max (preview)?
MiniMax M3 has the edge for coding in this comparison, averaging 72.2 versus 51. Inside this category, SWE-bench Pro is the benchmark that creates the most daylight between them.
Which is better for math, MiniMax M3 or Qwen 3.6 Max (preview)?
MiniMax M3 has the edge for math in this comparison, averaging 85.7 versus 18.4. Qwen 3.6 Max (preview) stays close enough that the answer can still flip depending on your workload.
Which is better for agentic tasks, MiniMax M3 or Qwen 3.6 Max (preview)?
MiniMax M3 has the edge for agentic tasks in this comparison, averaging 72.3 versus 65.4. Inside this category, τ²-bench results 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.