Skip to main content

Model comparison

Claude Opus 4.5 vs Qwen3.5 397B

Data verified

Head-to-head evidence from 48 shared benchmark results across 8 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

64.22/100
Margin
7.2pts
← winning
57.01/100
5 category wins3 category wins

Public leaderboard positions: Claude Opus 4.5 #34 (Supported); Qwen3.5 397B #71 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Claude Opus 4.5 and Qwen3.5 397B share 48 comparable benchmark results. 8 of 8 categories are comparable. 11 results are unique to Claude Opus 4.5; 7 to Qwen3.5 397B.

Updated July 22, 2026
Shared results
48
Claude Opus 4.5 only
11
Qwen3.5 397B only
7
Comparable categories
8 / 8

Pick Claude Opus 4.5 if you want the stronger benchmark profile. Qwen3.5 397B only becomes the better choice if mathematics is the priority or you want the cheaper token bill.

Why this result

Claude Opus 4.5 is clearly ahead on the BenchAlign aggregate, 64.22 to 57.01. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Claude Opus 4.5's sharpest advantage is in agentic, where it averages 62.6 against 56.5. The single biggest benchmark swing on the page is CharXiv, 68.5% to 80.8%. Qwen3.5 397B does hit back in mathematics, so the answer changes if that is the part of the workload you care about most.

Claude Opus 4.5 is also the more expensive model on tokens at $5.00 input / $25.00 output per 1M tokens, versus $0.60 input / $3.60 output per 1M tokens for Qwen3.5 397B. That is roughly 6.9x on output cost alone. Claude Opus 4.5 gives you the larger context window at 200K, 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 Claude Opus 4.5 and Qwen3.5 397B
CategoryClaude Opus 4.5ΔQwen3.5 397B
MathClaude Opus 4.557.5Margin 33.1Qwen3.5 397B90.6
Inst. FollowingClaude Opus 4.569.5Margin 23.1Qwen3.5 397B92.6
MultimodalClaude Opus 4.569.9Margin 9.7Qwen3.5 397B79.6
AgenticClaude Opus 4.562.6Margin 6.1Qwen3.5 397B56.5
CodingClaude Opus 4.571.7Margin 5.2Qwen3.5 397B66.5
KnowledgeClaude Opus 4.558.1Margin 1.5Qwen3.5 397B56.6
ReasoningClaude Opus 4.564.4Margin 1.2Qwen3.5 397B63.2
MultilingualClaude Opus 4.585.7Margin 1.0Qwen3.5 397B84.7

Decisive benchmark drivers

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

More
A · Claude Opus 4.5B · Qwen3.5 397B
  1. CharXiv

    Multimodal
    Source ↗
    A 68.5%B 80.8%
    Winner: Qwen3.5 397BΔ 12.3
    CharXiv: Claude Opus 4.5 scored 68.5%; Qwen3.5 397B scored 80.8%. Qwen3.5 397B wins this benchmark.
  2. MMMU-Pro

    Multimodal
    Source ↗
    A 70.6%B 79%
    Winner: Qwen3.5 397BΔ 8.4
    MMMU-Pro: Claude Opus 4.5 scored 70.6%; Qwen3.5 397B scored 79%. Qwen3.5 397B wins this benchmark.
  3. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 59.3%B 52.5%
    Winner: Claude Opus 4.5Δ 6.8
    Terminal-Bench 2.0: Claude Opus 4.5 scored 59.3%; Qwen3.5 397B scored 52.5%. Claude Opus 4.5 wins this benchmark.
  4. SWE-bench Pro

    Coding
    Source ↗
    A 57.1%B 50.9%
    Winner: Claude Opus 4.5Δ 6.2
    SWE-bench Pro: Claude Opus 4.5 scored 57.1%; Qwen3.5 397B scored 50.9%. Claude Opus 4.5 wins this benchmark.
  5. SWE-bench Verified

    Coding
    Source ↗
    A 80.9%B 76.2%
    Winner: Claude Opus 4.5Δ 4.7
    SWE-bench Verified: Claude Opus 4.5 scored 80.9%; Qwen3.5 397B scored 76.2%. Claude Opus 4.5 wins this benchmark.

Operational comparison

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

MetricClaude Opus 4.5Qwen3.5 397BComparison
Input / output priceUSD per 1M tokensClaude Opus 4.5$5 input / $25 outputQwen3.5 397B$0.6 input / $3.6 outputQwen3.5 397B has the lower combined listed price.
Generation speedtokens per secondClaude Opus 4.546 tok/sQwen3.5 397B96 tok/sQwen3.5 397B has the higher measured throughput.
First-answer latencyseconds to first tokenClaude Opus 4.51.01 sQwen3.5 397B2.44 sClaude Opus 4.5 reaches the first token sooner.
Context windowmaximum listed tokensClaude Opus 4.5200KQwen3.5 397B128KClaude Opus 4.5 lists the larger context window.

Benchmark Deep Dive

AgenticClaude Opus 4.5 wins
BenchmarkClaude Opus 4.5Qwen3.5 397BResult
Terminal-Bench 2.0Source 59.3%52.5%Claude Opus 4.5 leads
OSWorld-VerifiedSource 66.3%Not comparable
OSWorldSource 66.3%Not comparable
Claw-EvalSource 59.6%56.8%Claude Opus 4.5 leads
QwenClawBenchSource 52.3%51.8%Claude Opus 4.5 leads
τ³-bench resultsSource 70.2%68.4%Claude Opus 4.5 leads
VITA-BenchSource 23.3%43.7%Qwen3.5 397B leads
DeepPlanningSource 26.4%37.6%Qwen3.5 397B leads
ToolathlonSource 43.5%36.3%Claude Opus 4.5 leads
MCP AtlasSource 42.3%46.1%Qwen3.5 397B leads
MCP-TasksSource 71.8%74.2%Qwen3.5 397B leads
WideResearchSource 76.4%74.0%Claude Opus 4.5 leads
CyberGymSource 50.6%Not comparable
τ²-bench resultsSource 86.3%95.6%Qwen3.5 397B leads
Gert LabsSource 64.23%46.76%Claude Opus 4.5 leads
JobBenchSource 32.3%Not comparable
BrowseCompSource 62%Not comparable
ResearchClawBenchSource 14.2%Not comparable
AA Agentic IndexSource 19.9%Not comparable
APEX-Agents-AASource 15.3%Not comparable
GDPval-AASource 23.1%Not comparable
GDPval-AASource 962Not comparable
CodingClaude Opus 4.5 wins
BenchmarkClaude Opus 4.5Qwen3.5 397BResult
SWE-bench VerifiedSource 80.9%76.2%Claude Opus 4.5 leads
LiveCodeBench v6Source 84.8%83.6%Claude Opus 4.5 leads
SWE-bench ProSource 57.1%50.9%Claude Opus 4.5 leads
SWE MultilingualSource 77.5%Not comparable
NL2RepoSource 43.2%Not comparable
AA-SciCodeSource 47.0%42.0%Claude Opus 4.5 leads
AA Coding IndexSource 48.2%Not comparable
ReasoningClaude Opus 4.5 wins
BenchmarkClaude Opus 4.5Qwen3.5 397BResult
LongBench v2Source 64.4%63.2%Claude Opus 4.5 leads
AI-NeedleSource 74%68.7%Claude Opus 4.5 leads
AA-LCRSource 65.3%65.7%Qwen3.5 397B leads
CritPtSource 0.3%1.7%Qwen3.5 397B leads
KnowledgeClaude Opus 4.5 wins
BenchmarkClaude Opus 4.5Qwen3.5 397BResult
GPQASource 87%88.4%Qwen3.5 397B leads
SuperGPQASource 70.6%70.4%Claude Opus 4.5 leads
MMLU-ProSource 89.5%87.8%Claude Opus 4.5 leads
MMLU-ReduxSource 96.6%94.9%Claude Opus 4.5 leads
C-EvalSource 92.2%93%Qwen3.5 397B leads
HLESource 30.8%28.7%Claude Opus 4.5 leads
Artificial Analysis Intelligence IndexSource 34.7%33.7%Claude Opus 4.5 leads
AA-GPQA DiamondSource 81.0%89.3%Qwen3.5 397B leads
AA-HLESource 12.9%27.3%Qwen3.5 397B leads
AA-Omniscience IndexSource -3.9%-29.8%Claude Opus 4.5 leads
AA-Omniscience AccuracySource 40.7%31.4%Claude Opus 4.5 leads
AA-Omniscience Hallucination RateSource 75.4%89.1%Claude Opus 4.5 leads
AA MMLU-ProSource 88.9%Not comparable
MathQwen3.5 397B wins
BenchmarkClaude Opus 4.5Qwen3.5 397BResult
AIME26Source 95.1%93.3%Claude Opus 4.5 leads
HMMT Feb 2025Source 92.9%94.8%Qwen3.5 397B leads
HMMT Nov 2025Source 93.3%92.7%Claude Opus 4.5 leads
HMMT Feb 2026Source 85.3%87.9%Qwen3.5 397B leads
MMAnswerBenchSource 84.0%80.9%Claude Opus 4.5 leads
FrontierMath v2 (Tiers 1-3)Source 20.690%Not comparable
FrontierMath v2 (Tier 4)Source 4.167%Not comparable
MultilingualClaude Opus 4.5 wins
BenchmarkClaude Opus 4.5Qwen3.5 397BResult
MMLU-ProXSource 85.7%84.7%Claude Opus 4.5 leads
NOVA-63Source 56.7%59.1%Qwen3.5 397B leads
MultimodalQwen3.5 397B wins
BenchmarkClaude Opus 4.5Qwen3.5 397BResult
MMMU-ProSource 70.6%79%Qwen3.5 397B leads
MathVisionSource 74.3%88.6%Qwen3.5 397B leads
CharXivSource 68.5%80.8%Qwen3.5 397B leads
VideoMMMUSource 84.4%84.7%Qwen3.5 397B leads
ScreenSpot ProSource 45.7%65.6%Qwen3.5 397B leads
V*Source 67.0%95.8%Qwen3.5 397B leads
AA-MMMU-ProSource 71.2%77.3%Qwen3.5 397B leads
Design Arena WebsiteSource 1277Not comparable
Inst. FollowingQwen3.5 397B wins
BenchmarkClaude Opus 4.5Qwen3.5 397BResult
IFEvalSource 90.9%92.6%Qwen3.5 397B leads
IFBenchSource 58%Not comparable
AA-IFBenchSource 43.0%78.8%Qwen3.5 397B leads
Frequently Asked Questions (9)

Which is better, Claude Opus 4.5 or Qwen3.5 397B?

Claude Opus 4.5 is ahead on BenchLM's BenchAlign leaderboard, 64.22 to 57.01. The biggest single separator in this matchup is CharXiv, where the scores are 68.5% and 80.8%.

Which is better for knowledge tasks, Claude Opus 4.5 or Qwen3.5 397B?

Claude Opus 4.5 has the edge for knowledge tasks in this comparison, averaging 58.1 versus 56.6. Inside this category, AA-Omniscience Index is the benchmark that creates the most daylight between them.

Which is better for coding, Claude Opus 4.5 or Qwen3.5 397B?

Claude Opus 4.5 has the edge for coding in this comparison, averaging 71.7 versus 66.5. Inside this category, SWE-bench Pro is the benchmark that creates the most daylight between them.

Which is better for math, Claude Opus 4.5 or Qwen3.5 397B?

Qwen3.5 397B has the edge for math in this comparison, averaging 90.6 versus 57.5. Inside this category, MMAnswerBench is the benchmark that creates the most daylight between them.

Which is better for reasoning, Claude Opus 4.5 or Qwen3.5 397B?

Claude Opus 4.5 has the edge for reasoning in this comparison, averaging 64.4 versus 63.2. Inside this category, AI-Needle is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, Claude Opus 4.5 or Qwen3.5 397B?

Claude Opus 4.5 has the edge for agentic tasks in this comparison, averaging 62.6 versus 56.5. Inside this category, VITA-Bench is the benchmark that creates the most daylight between them.

Which is better for multimodal and grounded tasks, Claude Opus 4.5 or Qwen3.5 397B?

Qwen3.5 397B has the edge for multimodal and grounded tasks in this comparison, averaging 79.6 versus 69.9. Inside this category, V* is the benchmark that creates the most daylight between them.

Which is better for instruction following, Claude Opus 4.5 or Qwen3.5 397B?

Qwen3.5 397B has the edge for instruction following in this comparison, averaging 92.6 versus 69.5. Inside this category, AA-IFBench is the benchmark that creates the most daylight between them.

Which is better for multilingual tasks, Claude Opus 4.5 or Qwen3.5 397B?

Claude Opus 4.5 has the edge for multilingual tasks in this comparison, averaging 85.7 versus 84.7. Inside this category, NOVA-63 is the benchmark that creates the most daylight between them.

Related Comparisons

Last updated: July 22, 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.