Skip to main content

Model comparison

MiniMax M3 vs Qwen3.5 397B

Data verified

Head-to-head evidence from 26 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

MiniMax
69.8/100
Margin
13.1pts
← winning
56.73/100
2 category wins2 category wins

Verified leaderboard positions: MiniMax M3 #18; Qwen3.5 397B #20

BenchAlign evidence: MiniMax M3 supported; Qwen3.5 397B estimated. Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. MiniMax M3 and Qwen3.5 397B share 26 comparable benchmark results. 4 of 8 categories are comparable. 19 results are unique to MiniMax M3; 30 to Qwen3.5 397B.

Updated July 16, 2026
Shared results
26
MiniMax M3 only
19
Qwen3.5 397B only
30
Comparable categories
4 / 8

Pick MiniMax M3 if you want the stronger benchmark profile. Qwen3.5 397B only becomes the better choice if multimodal & grounded is the priority.

Confidence note. This is a partial-evidence comparison with 26 shared benchmark results across 6 evidence categories; 4 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 is clearly ahead on the provisional aggregate, 70 to 59. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

MiniMax M3's sharpest advantage is in agentic, where it averages 72.3 against 56.5. The single biggest benchmark swing on the page is BrowseComp, 83.5% to 62%. Qwen3.5 397B does hit back in multimodal & grounded, so the answer changes if that is the part of the workload you care about most.

Qwen3.5 397B is also the more expensive model on tokens at $0.60 input / $3.60 output per 1M tokens, versus $0.30 input / $1.20 output per 1M tokens for MiniMax M3. That is roughly 3.0x on output cost alone. MiniMax M3 gives you the larger context window at 1M, compared with 128K for Qwen3.5 397B.

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 scores and score margins for MiniMax M3 and Qwen3.5 397B
CategoryMiniMax M3ΔQwen3.5 397B
AgenticMiniMax M372.3Margin 15.8Qwen3.5 397B56.5
MultimodalMiniMax M364.9Margin 14.7Qwen3.5 397B79.6
CodingMiniMax M372.2Margin 5.7Qwen3.5 397B66.5
MathMiniMax M385.7Margin 4.9Qwen3.5 397B90.6
ReasoningMiniMax M3Not measuredMarginNo overlapQwen3.5 397B63.2
KnowledgeMiniMax M3Not measuredMarginNo overlapQwen3.5 397B56.9
MultilingualMiniMax M3Not measuredMarginNo overlapQwen3.5 397B84.7
Inst. FollowingMiniMax M3Not measuredMarginNo overlapQwen3.5 397B92.6

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · MiniMax M3B · Qwen3.5 397B
  1. BrowseComp

    Agentic
    Source ↗
    A 83.5%B 62%
    Winner: MiniMax M3Δ 21.5
    BrowseComp: MiniMax M3 scored 83.5%; Qwen3.5 397B scored 62%. MiniMax M3 wins this benchmark.
  2. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 66%B 52.5%
    Winner: MiniMax M3Δ 13.5
    Terminal-Bench 2.0: MiniMax M3 scored 66%; Qwen3.5 397B scored 52.5%. MiniMax M3 wins this benchmark.
  3. SWE-bench Pro

    Coding
    Source ↗
    A 59%B 50.9%
    Winner: MiniMax M3Δ 8.1
    SWE-bench Pro: MiniMax M3 scored 59%; Qwen3.5 397B scored 50.9%. MiniMax M3 wins this benchmark.
  4. SWE-bench Verified

    Coding
    Source ↗
    A 80.5%B 76.2%
    Winner: MiniMax M3Δ 4.3
    SWE-bench Verified: MiniMax M3 scored 80.5%; Qwen3.5 397B scored 76.2%. MiniMax M3 wins this benchmark.
  5. MMMU-Pro

    Multimodal
    Source ↗
    A 78.1%B 79%
    Winner: Qwen3.5 397BΔ 0.9
    MMMU-Pro: MiniMax M3 scored 78.1%; Qwen3.5 397B scored 79%. Qwen3.5 397B wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricMiniMax M3Qwen3.5 397BComparison
Input / output priceUSD per 1M tokensMiniMax M3$0.3 input / $1.2 outputQwen3.5 397B$0.6 input / $3.6 outputMiniMax M3 has the lower combined listed price.
Generation speedtokens per secondMiniMax M3Not availableQwen3.5 397B96 tok/sA complete speed comparison is not available.
First-answer latencyseconds to first tokenMiniMax M3Not availableQwen3.5 397B2.44 sA complete latency comparison is not available.
Context windowmaximum listed tokensMiniMax M31MQwen3.5 397B128KMiniMax M3 lists the larger context window.

Benchmark Deep Dive

AgenticMiniMax M3 wins
BenchmarkMiniMax M3Qwen3.5 397BResult
Terminal-Bench 2.0Source 66%52.5%MiniMax M3 leads
BrowseCompSource 83.5%62%MiniMax M3 leads
OSWorld-VerifiedSource 70.1%Not comparable
MCP AtlasSource 74.2%46.1%MiniMax M3 leads
Claw-EvalSource 74.5%56.8%MiniMax M3 leads
AA Agentic IndexSource 35.4%19.9%MiniMax M3 leads
τ²-bench resultsSource 88.9%95.6%Qwen3.5 397B leads
GDPval-AASource 44.7%23.1%MiniMax M3 leads
GDPval-AASource 1395962MiniMax M3 leads
GDPval rubricsSource 74.7%Not comparable
BankerToolBenchSource 76.1%Not comparable
ResearchClawBenchSource 19.8%14.2%MiniMax M3 leads
OSWorld 2.0Source 4.6%Not comparable
AA BriefcaseSource 1110Not comparable
AA EnterpriseOps-GymSource 32.1%Not comparable
AA Harvey LABSource 6.7%Not comparable
QwenClawBenchSource 51.8%Not comparable
τ³-bench resultsSource 68.4%Not comparable
VITA-BenchSource 43.7%Not comparable
DeepPlanningSource 37.6%Not comparable
ToolathlonSource 36.3%Not comparable
MCP-TasksSource 74.2%Not comparable
WideResearchSource 74.0%Not comparable
Gert LabsSource 46.76%Not comparable
APEX-Agents-AASource 15.3%Not comparable
CodingMiniMax M3 wins
BenchmarkMiniMax M3Qwen3.5 397BResult
SWE-bench VerifiedSource 80.5%76.2%MiniMax M3 leads
SWE-bench ProSource 59%50.9%MiniMax M3 leads
Terminal-Bench 2.0Source 66.0%Not comparable
NL2RepoSource 42.1%Not comparable
AA Coding IndexSource 58.6%48.2%MiniMax M3 leads
Terminal-Bench HardSource 42.4%40.9%MiniMax M3 leads
AA-SciCodeSource 45.4%42.0%MiniMax M3 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
LiveCodeBench v6Source 83.6%Not comparable
Reasoning
BenchmarkMiniMax M3Qwen3.5 397BResult
AA-LCRSource 74.0%65.7%MiniMax M3 leads
CritPtSource 3.7%1.7%MiniMax M3 leads
LongBench v2Source 63.2%Not comparable
AI-NeedleSource 68.7%Not comparable
Knowledge
BenchmarkMiniMax M3Qwen3.5 397BResult
Artificial Analysis Intelligence IndexSource 44.4%33.7%MiniMax M3 leads
AA-GPQA DiamondSource 92.9%89.3%MiniMax M3 leads
AA-HLESource 37.1%27.3%MiniMax M3 leads
AA-Omniscience IndexSource 1.4%-29.8%MiniMax M3 leads
AA-Omniscience AccuracySource 15.0%31.4%Qwen3.5 397B leads
AA-Omniscience Hallucination RateSource 16.1%89.1%MiniMax M3 leads
AA Openness IndexSource 33.3%Not comparable
GPQASource 88.4%Not comparable
SuperGPQASource 70.4%Not comparable
MMLU-ProSource 87.8%Not comparable
MMLU-ReduxSource 94.9%Not comparable
C-EvalSource 93%Not comparable
HLESource 28.7%Not comparable
MathQwen3.5 397B wins
BenchmarkMiniMax M3Qwen3.5 397BResult
USAMO 2026Source 85.7%Not comparable
AIME26Source 93.3%Not comparable
HMMT Feb 2025Source 94.8%Not comparable
HMMT Nov 2025Source 92.7%Not comparable
HMMT Feb 2026Source 87.9%Not comparable
MMAnswerBenchSource 80.9%Not comparable
Multilingual
BenchmarkMiniMax M3Qwen3.5 397BResult
MMLU-ProXSource 84.7%Not comparable
NOVA-63Source 59.1%Not comparable
MultimodalQwen3.5 397B wins
BenchmarkMiniMax M3Qwen3.5 397BResult
OfficeQA ProSource 45.1%Not comparable
OmniDocBench 1.5Source 91.6%Not comparable
MMMU-ProSource 78.1%79%Qwen3.5 397B leads
VideoMMMUSource 84.6%84.7%Qwen3.5 397B leads
Video-MME (with subtitle)Source 85.4%Not comparable
Design Arena WebsiteSource 1294Not comparable
AA-MMMU-ProSource 78.6%77.3%MiniMax M3 leads
MathVisionSource 88.6%Not comparable
CharXivSource 80.8%Not comparable
ScreenSpot ProSource 65.6%Not comparable
V*Source 95.8%Not comparable
Inst. Following
BenchmarkMiniMax M3Qwen3.5 397BResult
AA-IFBenchSource 82.9%78.8%MiniMax M3 leads
IFEvalSource 92.6%Not comparable
Frequently Asked Questions (5)

Which is better, MiniMax M3 or Qwen3.5 397B?

MiniMax M3 is ahead on BenchLM's provisional leaderboard, 70 to 59. The biggest single separator in this matchup is BrowseComp, where the scores are 83.5% and 62%.

Which is better for coding, MiniMax M3 or Qwen3.5 397B?

MiniMax M3 has the edge for coding in this comparison, averaging 72.2 versus 66.5. Inside this category, AA Coding Index is the benchmark that creates the most daylight between them.

Which is better for math, MiniMax M3 or Qwen3.5 397B?

Qwen3.5 397B has the edge for math in this comparison, averaging 90.6 versus 85.7. MiniMax M3 stays close enough that the answer can still flip depending on your workload.

Which is better for agentic tasks, MiniMax M3 or Qwen3.5 397B?

MiniMax M3 has the edge for agentic tasks in this comparison, averaging 72.3 versus 56.5. Inside this category, GDPval-AA is the benchmark that creates the most daylight between them.

Which is better for multimodal and grounded tasks, MiniMax M3 or Qwen3.5 397B?

Qwen3.5 397B has the edge for multimodal and grounded tasks in this comparison, averaging 79.6 versus 64.9. Inside this category, AA-MMMU-Pro is the benchmark that creates the most daylight between them.

Related Comparisons

Last updated: July 16, 2026

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.