Skip to main content

Model comparison

Llama 4 Maverick vs Qwen3.5 397B

Data verified

Head-to-head evidence from 16 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

23.49/100
Margin
33.5pts
winning →
57.01/100
0 category wins1 category wins

Public leaderboard positions: Llama 4 Maverick #191 (Supported); Qwen3.5 397B #71 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Llama 4 Maverick and Qwen3.5 397B share 16 comparable benchmark results. 1 of 8 categories are comparable. 2 results are unique to Llama 4 Maverick; 39 to Qwen3.5 397B.

Updated July 20, 2026
Shared results
16
Llama 4 Maverick only
2
Qwen3.5 397B only
39
Comparable categories
1 / 8

Pick Qwen3.5 397B if you want the stronger benchmark profile. Llama 4 Maverick only becomes the better choice if you want the cheaper token bill or you need the larger 1M context window.

Confidence note. This is a partial-evidence comparison with 16 shared benchmark results across 6 evidence categories; 1 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Qwen3.5 397B is clearly ahead on the BenchAlign aggregate, 57.01 to 23.49. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Qwen3.5 397B's sharpest advantage is in mathematics, where it averages 90.6 against 0.7.

Qwen3.5 397B is also the more expensive model on tokens at $0.60 input / $3.60 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Llama 4 Maverick. That is roughly Infinityx on output cost alone. Llama 4 Maverick 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 Llama 4 Maverick and Qwen3.5 397B
CategoryLlama 4 MaverickΔQwen3.5 397B
MathLlama 4 Maverick0.7Margin 89.9Qwen3.5 397B90.6
AgenticLlama 4 MaverickNot measuredMarginNo overlapQwen3.5 397B56.5
CodingLlama 4 MaverickNot measuredMarginNo overlapQwen3.5 397B66.5
ReasoningLlama 4 MaverickNot measuredMarginNo overlapQwen3.5 397B63.2
KnowledgeLlama 4 MaverickNot measuredMarginNo overlapQwen3.5 397B56.6
MultilingualLlama 4 MaverickNot measuredMarginNo overlapQwen3.5 397B84.7
MultimodalLlama 4 MaverickNot measuredMarginNo overlapQwen3.5 397B79.6
Inst. FollowingLlama 4 MaverickNot measuredMarginNo overlapQwen3.5 397B92.6

Operational comparison

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

MetricLlama 4 MaverickQwen3.5 397BComparison
Input / output priceUSD per 1M tokensLlama 4 Maverick$0 input / $0 outputQwen3.5 397B$0.6 input / $3.6 outputLlama 4 Maverick has the lower combined listed price.
Generation speedtokens per secondLlama 4 Maverick121 tok/sQwen3.5 397B96 tok/sLlama 4 Maverick has the higher measured throughput.
First-answer latencyseconds to first tokenLlama 4 Maverick0.95 sQwen3.5 397B2.44 sLlama 4 Maverick reaches the first token sooner.
Context windowmaximum listed tokensLlama 4 Maverick1MQwen3.5 397B128KLlama 4 Maverick lists the larger context window.

Benchmark Deep Dive

Agentic
BenchmarkLlama 4 MaverickQwen3.5 397BResult
AA Agentic IndexSource 1.3%19.9%Qwen3.5 397B leads
τ²-bench resultsSource 17.8%95.6%Qwen3.5 397B leads
GDPval-AASource 0.0%23.1%Qwen3.5 397B leads
GDPval-AASource -16962Qwen3.5 397B leads
Terminal-Bench 2.0Source 52.5%Not comparable
BrowseCompSource 62%Not comparable
Claw-EvalSource 56.8%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 AtlasSource 46.1%Not comparable
MCP-TasksSource 74.2%Not comparable
WideResearchSource 74.0%Not comparable
Gert LabsSource 46.76%Not comparable
ResearchClawBenchSource 14.2%Not comparable
APEX-Agents-AASource 15.3%Not comparable
Coding
BenchmarkLlama 4 MaverickQwen3.5 397BResult
AA Coding IndexSource 16.3%48.2%Qwen3.5 397B leads
AA-SciCodeSource 33.1%42.0%Qwen3.5 397B leads
SWE-bench VerifiedSource 76.2%Not comparable
LiveCodeBench v6Source 83.6%Not comparable
SWE-bench ProSource 50.9%Not comparable
Reasoning
BenchmarkLlama 4 MaverickQwen3.5 397BResult
AA-LCRSource 46.0%65.7%Qwen3.5 397B leads
CritPtSource 0.0%1.7%Qwen3.5 397B leads
LongBench v2Source 63.2%Not comparable
AI-NeedleSource 68.7%Not comparable
Knowledge
BenchmarkLlama 4 MaverickQwen3.5 397BResult
Artificial Analysis Intelligence IndexSource 14.3%33.7%Qwen3.5 397B leads
AA-GPQA DiamondSource 67.1%89.3%Qwen3.5 397B leads
AA-HLESource 4.8%27.3%Qwen3.5 397B leads
AA-Omniscience IndexSource -41.8%-29.8%Qwen3.5 397B leads
AA-Omniscience AccuracySource 24.3%31.4%Qwen3.5 397B leads
AA-Omniscience Hallucination RateSource 87.3%89.1%Llama 4 Maverick leads
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
BenchmarkLlama 4 MaverickQwen3.5 397BResult
FrontierMath v2 (Tiers 1-3)Source 0.690%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
BenchmarkLlama 4 MaverickQwen3.5 397BResult
MMLU-ProXSource 84.7%Not comparable
NOVA-63Source 59.1%Not comparable
Multimodal
BenchmarkLlama 4 MaverickQwen3.5 397BResult
AA-MMMU-ProSource 62.1%77.3%Qwen3.5 397B leads
Design Arena WebsiteSource 903Not 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
Inst. Following
BenchmarkLlama 4 MaverickQwen3.5 397BResult
AA-IFBenchSource 43.0%78.8%Qwen3.5 397B leads
IFEvalSource 92.6%Not comparable
Frequently Asked Questions (2)

Which is better, Llama 4 Maverick or Qwen3.5 397B?

Qwen3.5 397B is ahead on BenchLM's BenchAlign leaderboard, 57.01 to 23.49.

Which is better for math, Llama 4 Maverick or Qwen3.5 397B?

Qwen3.5 397B has the edge for math in this comparison, averaging 90.6 versus 0.7. Llama 4 Maverick stays close enough that the answer can still flip depending on your workload.

Self-host vs API cost

Estimates at 50,000 req/day · 1000 tokens/req average.

Llama 4 Maverick
API / mo$0
Self-host / mo$2,610
Break-even
Qwen3.5 397B
API / mo$3,150
Self-host / moNot listed
Break-even
Proprietary model — self-hosting not applicable.
Model the full break-even

Related Comparisons

Last updated: July 20, 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.