Skip to main content

Model comparison

MiMo-V2.5-Pro vs Qwen3.5 397B

Data verified

Head-to-head evidence from 22 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

70.19/100
Margin
13.2pts
← winning
57.01/100
1 category wins2 category wins

Public leaderboard positions: MiMo-V2.5-Pro #14 (Supported); Qwen3.5 397B #71 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. MiMo-V2.5-Pro and Qwen3.5 397B share 22 comparable benchmark results. 3 of 8 categories are comparable. 9 results are unique to MiMo-V2.5-Pro; 33 to Qwen3.5 397B.

Updated July 21, 2026
Shared results
22
MiMo-V2.5-Pro only
9
Qwen3.5 397B only
33
Comparable categories
3 / 8

Pick MiMo-V2.5-Pro if you want the stronger benchmark profile. Qwen3.5 397B 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 22 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

MiMo-V2.5-Pro is clearly ahead on the BenchAlign aggregate, 70.19 to 57.01. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

MiMo-V2.5-Pro's sharpest advantage is in agentic, where it averages 68.4 against 56.5. The single biggest benchmark swing on the page is HLE, 48% to 28.7%. Qwen3.5 397B does hit back in coding, so the answer changes if that is the part of the workload you care about most.

MiMo-V2.5-Pro is the reasoning model in the pair, while Qwen3.5 397B 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. MiMo-V2.5-Pro 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 MiMo-V2.5-Pro and Qwen3.5 397B
CategoryMiMo-V2.5-ProΔQwen3.5 397B
AgenticMiMo-V2.5-Pro68.4Margin 11.9Qwen3.5 397B56.5
CodingMiMo-V2.5-Pro57.2Margin 9.3Qwen3.5 397B66.5
KnowledgeMiMo-V2.5-Pro48.0Margin 8.6Qwen3.5 397B56.6
ReasoningMiMo-V2.5-ProNot measuredMarginNo overlapQwen3.5 397B63.2
MathMiMo-V2.5-ProNot measuredMarginNo overlapQwen3.5 397B90.6
MultilingualMiMo-V2.5-ProNot measuredMarginNo overlapQwen3.5 397B84.7
MultimodalMiMo-V2.5-ProNot measuredMarginNo overlapQwen3.5 397B79.6
Inst. FollowingMiMo-V2.5-ProNot measuredMarginNo overlapQwen3.5 397B92.6

Decisive benchmark drivers

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

More
A · MiMo-V2.5-ProB · Qwen3.5 397B
  1. HLE

    Knowledge
    Source ↗
    A 48%B 28.7%
    Winner: MiMo-V2.5-ProΔ 19.3
    HLE: MiMo-V2.5-Pro scored 48%; Qwen3.5 397B scored 28.7%. MiMo-V2.5-Pro wins this benchmark.
  2. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 68.4%B 52.5%
    Winner: MiMo-V2.5-ProΔ 15.9
    Terminal-Bench 2.0: MiMo-V2.5-Pro scored 68.4%; Qwen3.5 397B scored 52.5%. MiMo-V2.5-Pro wins this benchmark.
  3. SWE-bench Pro

    Coding
    Source ↗
    A 57.2%B 50.9%
    Winner: MiMo-V2.5-ProΔ 6.3
    SWE-bench Pro: MiMo-V2.5-Pro scored 57.2%; Qwen3.5 397B scored 50.9%. MiMo-V2.5-Pro wins this benchmark.

Operational comparison

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

MetricMiMo-V2.5-ProQwen3.5 397BComparison
Input / output priceUSD per 1M tokensMiMo-V2.5-ProNot availableQwen3.5 397B$0.6 input / $3.6 outputA complete price comparison is not available.
Generation speedtokens per secondMiMo-V2.5-ProNot availableQwen3.5 397B96 tok/sA complete speed comparison is not available.
First-answer latencyseconds to first tokenMiMo-V2.5-ProNot availableQwen3.5 397B2.44 sA complete latency comparison is not available.
Context windowmaximum listed tokensMiMo-V2.5-Pro1MQwen3.5 397B128KMiMo-V2.5-Pro lists the larger context window.

Benchmark Deep Dive

AgenticMiMo-V2.5-Pro wins
BenchmarkMiMo-V2.5-ProQwen3.5 397BResult
Claw-EvalSource 63.8%56.8%MiMo-V2.5-Pro leads
GDPval-AASource 1265962MiMo-V2.5-Pro leads
τ³-bench resultsSource 72.9%68.4%MiMo-V2.5-Pro leads
Terminal-Bench 2.0Source 68.4%52.5%MiMo-V2.5-Pro leads
AA Agentic IndexSource 29.1%19.9%MiMo-V2.5-Pro leads
τ²-bench resultsSource 94.2%95.6%Qwen3.5 397B leads
GDPval-AASource 38.3%23.1%MiMo-V2.5-Pro leads
APEX-Agents-AASource 2.4%15.3%Qwen3.5 397B leads
Gert LabsSource 62.70%46.76%MiMo-V2.5-Pro leads
AA BriefcaseSource 873Not comparable
AA ITBenchSource 38.2%Not comparable
terminalBenchHardSource 43.2%Not comparable
aaTerminalBench21Source 65.2%Not comparable
AA Harvey LABSource 73.3%Not comparable
BrowseCompSource 62%Not comparable
QwenClawBenchSource 51.8%Not comparable
VITA-BenchSource 43.7%Not comparable
DeepPlanningSource 37.6%Not comparable
ToolathlonSource 36.3%Not comparable
MCP AtlasSource 46.1%Not comparable
MCP-TasksSource 74.2%Not comparable
WideResearchSource 74.0%Not comparable
ResearchClawBenchSource 14.2%Not comparable
CodingQwen3.5 397B wins
BenchmarkMiMo-V2.5-ProQwen3.5 397BResult
SWE-bench ProSource 57.2%50.9%MiMo-V2.5-Pro leads
Terminal-Bench 2.0Source 68.4%Not comparable
AA Coding IndexSource 60.2%48.2%MiMo-V2.5-Pro leads
AA-SciCodeSource 50.2%42.0%MiMo-V2.5-Pro leads
SWE-bench VerifiedSource 76.2%Not comparable
LiveCodeBench v6Source 83.6%Not comparable
Reasoning
BenchmarkMiMo-V2.5-ProQwen3.5 397BResult
AA-LCRSource 73.3%65.7%MiMo-V2.5-Pro leads
CritPtSource 4.0%1.7%MiMo-V2.5-Pro leads
LongBench v2Source 63.2%Not comparable
AI-NeedleSource 68.7%Not comparable
KnowledgeQwen3.5 397B wins
BenchmarkMiMo-V2.5-ProQwen3.5 397BResult
HLESource 48%28.7%MiMo-V2.5-Pro leads
HLE w/o toolsSource 34%Not comparable
Artificial Analysis Intelligence IndexSource 42.2%33.7%MiMo-V2.5-Pro leads
AA-GPQA DiamondSource 86.6%89.3%Qwen3.5 397B leads
AA-HLESource 33.8%27.3%MiMo-V2.5-Pro leads
AA-Omniscience IndexSource 3.6%-29.8%MiMo-V2.5-Pro leads
AA-Omniscience AccuracySource 22.6%31.4%Qwen3.5 397B leads
AA-Omniscience Hallucination RateSource 24.5%89.1%MiMo-V2.5-Pro leads
AA Openness IndexSource 38.9%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
Math
BenchmarkMiMo-V2.5-ProQwen3.5 397BResult
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
BenchmarkMiMo-V2.5-ProQwen3.5 397BResult
MMLU-ProXSource 84.7%Not comparable
NOVA-63Source 59.1%Not comparable
Multimodal
BenchmarkMiMo-V2.5-ProQwen3.5 397BResult
Design Arena WebsiteSource 1298Not comparable
MMMU-ProSource 79%Not comparable
MathVisionSource 88.6%Not comparable
CharXivSource 80.8%Not comparable
VideoMMMUSource 84.7%Not comparable
ScreenSpot ProSource 65.6%Not comparable
V*Source 95.8%Not comparable
AA-MMMU-ProSource 77.3%Not comparable
Inst. Following
BenchmarkMiMo-V2.5-ProQwen3.5 397BResult
AA-IFBenchSource 79.9%78.8%MiMo-V2.5-Pro leads
IFEvalSource 92.6%Not comparable
Frequently Asked Questions (4)

Which is better, MiMo-V2.5-Pro or Qwen3.5 397B?

MiMo-V2.5-Pro is ahead on BenchLM's BenchAlign leaderboard, 70.19 to 57.01. The biggest single separator in this matchup is HLE, where the scores are 48% and 28.7%.

Which is better for knowledge tasks, MiMo-V2.5-Pro or Qwen3.5 397B?

Qwen3.5 397B has the edge for knowledge tasks in this comparison, averaging 56.6 versus 48. Inside this category, AA-Omniscience Hallucination Rate is the benchmark that creates the most daylight between them.

Which is better for coding, MiMo-V2.5-Pro or Qwen3.5 397B?

Qwen3.5 397B has the edge for coding in this comparison, averaging 66.5 versus 57.2. Inside this category, AA Coding Index is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, MiMo-V2.5-Pro or Qwen3.5 397B?

MiMo-V2.5-Pro has the edge for agentic tasks in this comparison, averaging 68.4 versus 56.5. Inside this category, GDPval-AA is the benchmark that creates the most daylight between them.

Related Comparisons

Last updated: July 21, 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.