LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

Llama 4 Scout vs Qwen3.5-Flash

Updated October 21, 2026. 45 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace Llama 4 Scout Replace Qwen3.5-Flash
Llama 4 ScoutMeta Llama
Coding
9.1
Overall
6.5
Output $/M
$0.30
0
Qwen3.5-FlashQwen
Coding
6.2
Overall
No data
Output $/M
$0.26
45
vs boards won of 45 shared

Decision reading

Pick Llama 4 Scout for coding: it scores 9.1 on the Vibe Coding Index against 6.2 for Qwen3.5-Flash.

Pick Qwen3.5-Flash when output tokens cost 1.2x less, or when a solved task costs $0.40 against $1.276.

Model ALlama 4 ScoutMeta Llama9.1/1001 independent benchmark$1.276 per solved taskModel BQwen3.5-FlashQwen6.2/1001 independent benchmark$0.40 per solved task
Shared results
45
Llama 4 Scout only
35
Qwen3.5-Flash only
10
Like-for-like categories
6 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
Llama 4 Scout 9.1 from 1 benchmark, Qwen3.5-Flash 6.2 from 1 benchmark.
Widest category
Math, 47.9 points apart on average over 4 shared benchmarks.
Widest single gap
OTIS Mock AIME: Llama 4 Scout 7.8% against Qwen3.5-Flash 84.4%, 76.7 points apart on a 0 to 100 scale.
Output price
Llama 4 Scout $0.30 against Qwen3.5-Flash $0.26 per million output tokens, 1.2x apart.
Context window
Llama 4 Scout takes 1.31M tokens, Qwen3.5-Flash 1M.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryLlama 4 ScoutBarsQwen3.5-FlashTies
Coding03
Intelligence01
Reasoning06
Math04
Knowledge010
General021
All shared boards045
CodingLike-for-likeFour of these boards feed the Vibe Coding Index03
Coding benchmarks for Llama 4 Scout and Qwen3.5-Flash
BenchmarkLlama 4 ScoutQwen3.5-FlashLeader
SWE-bench Verified9.1%Not measuredOnly Llama 4 Scout measured
LMArena WebDevNot measured1244Only Qwen3.5-Flash measured
AA Coding Index8.2Not measuredOnly Llama 4 Scout measured
LiveCodeBench (Vals AI)38.5%83.3%Qwen3.5-Flash
SWE-bench Verified (Vals AI)Not measured64.4%Only Qwen3.5-Flash measured
LiveCodeBench (Artificial Analysis run)29.9%Not measuredOnly Llama 4 Scout measured
SciCode21.3%Not measuredOnly Llama 4 Scout measured
Terminal-Bench Hard (Artificial Analysis run)1.5%Not measuredOnly Llama 4 Scout measured
ALE-BenchNot measured265.9Only Qwen3.5-Flash measured
LMArena Text: coding12861412Qwen3.5-Flash
LMArena Text: software and IT12861416Qwen3.5-Flash
LMArena WebDev: ReactNot measured1229Only Qwen3.5-Flash measured
LMArena WebDev: HTMLNot measured1291Only Qwen3.5-Flash measured
Neither model has a result on the other 25 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for Llama 4 Scout and Qwen3.5-Flash
BenchmarkLlama 4 ScoutQwen3.5-FlashLeader
AA Intelligence Index6.5Not measuredOnly Llama 4 Scout measured
Epoch Capabilities Index129.6143.5Qwen3.5-Flash
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
Agentic work benchmarks for Llama 4 Scout and Qwen3.5-Flash
BenchmarkLlama 4 ScoutQwen3.5-FlashLeader
AA Agentic Index0.5Not measuredOnly Llama 4 Scout measured
τ²-Bench Telecom15.5%Not measuredOnly Llama 4 Scout measured
Neither model has a result on the other 37 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index06
Reasoning benchmarks for Llama 4 Scout and Qwen3.5-Flash
BenchmarkLlama 4 ScoutQwen3.5-FlashLeader
GPQA Diamond47.0%82.8%Qwen3.5-Flash
MMMU Pro (Vals AI)58.8%81.9%Qwen3.5-Flash
GPQA Diamond (Artificial Analysis run)58.7%Not measuredOnly Llama 4 Scout measured
Humanity's Last Exam (Artificial Analysis run)3.8%Not measuredOnly Llama 4 Scout measured
GPQA Diamond (Epoch AI run)51.8%82.3%Qwen3.5-Flash
ARC-AGI0.5%Not measuredOnly Llama 4 Scout measured
ARC-AGI-20.0%Not measuredOnly Llama 4 Scout measured
Chess puzzlesNot measured14.0%Only Qwen3.5-Flash measured
Mystery game puzzlesNot measured20.0%Only Qwen3.5-Flash measured
LMArena Text: hard prompts12661403Qwen3.5-Flash
LMArena Text: expert12351407Qwen3.5-Flash
LMArena Text: hard prompts in English12931409Qwen3.5-Flash
LMArena Vision: diagrams1125Not measuredOnly Llama 4 Scout measured
LMArena Vision: homework1148Not measuredOnly Llama 4 Scout measured
Neither model has a result on the other 7 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index04
Math benchmarks for Llama 4 Scout and Qwen3.5-Flash
BenchmarkLlama 4 ScoutQwen3.5-FlashLeader
AIME19.0%92.5%Qwen3.5-Flash
MATH 500 (Vals AI)79.2%Not measuredOnly Llama 4 Scout measured
MGSM (Vals AI)88.0%Not measuredOnly Llama 4 Scout measured
AA Math Index14.0Not measuredOnly Llama 4 Scout measured
AIME 2025 (Artificial Analysis run)14.0%Not measuredOnly Llama 4 Scout measured
OTIS Mock AIME7.8%84.4%Qwen3.5-Flash
FrontierMath, tiers 1 to 3Not measured18.2%Only Qwen3.5-Flash measured
MATH level 562.3%Not measuredOnly Llama 4 Scout measured
LMArena Text: math12871407Qwen3.5-Flash
LMArena Text: mathematical work12861413Qwen3.5-Flash
Neither model has a result on the other 3 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index010
Knowledge benchmarks for Llama 4 Scout and Qwen3.5-Flash
BenchmarkLlama 4 ScoutQwen3.5-FlashLeader
MMLU Pro69.6%84.1%Qwen3.5-Flash
LegalBench (Vals AI)72.0%84.3%Qwen3.5-Flash
CaseLaw (Vals AI)Not measured56.0%Only Qwen3.5-Flash measured
CorpFin (Vals AI)46.8%63.6%Qwen3.5-Flash
TaxEval (Vals AI)55.2%72.2%Qwen3.5-Flash
MortgageTax (Vals AI)57.8%67.4%Qwen3.5-Flash
MedQA (Vals AI)50.9%Not measuredOnly Llama 4 Scout measured
MedCode (Vals AI)23.3%33.0%Qwen3.5-Flash
MMLU Pro (Artificial Analysis run)75.2%Not measuredOnly Llama 4 Scout measured
SimpleQA VerifiedNot measured20.3%Only Qwen3.5-Flash measured
LMArena Text: science12781415Qwen3.5-Flash
LMArena Text: medicine and healthcare12751407Qwen3.5-Flash
LMArena Text: legal and government12851398Qwen3.5-Flash
LMArena Text: business and finance12571394Qwen3.5-Flash
Neither model has a result on the other 3 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index5 boards, none shared
Web design benchmarks for Llama 4 Scout and Qwen3.5-Flash
BenchmarkLlama 4 ScoutQwen3.5-FlashLeader
Design Arena: websites750Not measuredOnly Llama 4 Scout measured
Design Arena: UI components767Not measuredOnly Llama 4 Scout measured
Design Arena: game dev773Not measuredOnly Llama 4 Scout measured
Design Arena: data viz896Not measuredOnly Llama 4 Scout measured
Design Arena: all code categories790Not measuredOnly Llama 4 Scout measured
Neither model has a result on the other 4 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index021
General benchmarks for Llama 4 Scout and Qwen3.5-Flash
BenchmarkLlama 4 ScoutQwen3.5-FlashLeader
SAGE (Vals AI)34.8%42.5%Qwen3.5-Flash
MedScribe (Vals AI)50.6%70.6%Qwen3.5-Flash
IFBench39.5%Not measuredOnly Llama 4 Scout measured
Long context reasoning (AA-LCR)27.7%Not measuredOnly Llama 4 Scout measured
DTBench57.9%82.9%Qwen3.5-Flash
LMCA12.0%29.1%Qwen3.5-Flash
Fiction.LiveBench36.0%Not measuredOnly Llama 4 Scout measured
LMArena Text12791397Qwen3.5-Flash
LMArena Text: instruction following12481374Qwen3.5-Flash
LMArena Text: creative writing12491343Qwen3.5-Flash
LMArena Vision1118Not measuredOnly Llama 4 Scout measured
LMArena Text: English13041404Qwen3.5-Flash
LMArena Text: non-English12521385Qwen3.5-Flash
LMArena Text: Chinese12551445Qwen3.5-Flash
LMArena Text: German12731390Qwen3.5-Flash
LMArena Text: Japanese12061369Qwen3.5-Flash
LMArena Text: Korean12071344Qwen3.5-Flash
LMArena Text: French12831413Qwen3.5-Flash
LMArena Text: Spanish12781400Qwen3.5-Flash
LMArena Text: Russian12631380Qwen3.5-Flash
LMArena Text: Polish12651376Qwen3.5-Flash
LMArena Text: multi-turn12801393Qwen3.5-Flash
LMArena Text: longer queries12651392Qwen3.5-Flash
LMArena Text: writing and language12691369Qwen3.5-Flash
LMArena Text: entertainment and media12421347Qwen3.5-Flash
LMArena Vision: English1143Not measuredOnly Llama 4 Scout measured
LMArena Vision: Chinese1106Not measuredOnly Llama 4 Scout measured
LMArena Vision: reading text1117Not measuredOnly Llama 4 Scout measured
LMArena Vision: humor1102Not measuredOnly Llama 4 Scout measured
LMArena Vision: creative writing1126Not measuredOnly Llama 4 Scout measured
Neither model has a result on the other 16 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Llama 4 Scout and Qwen3.5-Flash
BenchmarkLlama 4 ScoutQwen3.5-FlashLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

Llama 4 Scout$0.0003Fits in one request
Qwen3.5-Flash$0.0002Fits in one request

Qwen3.5-Flash costs less

Repository review

50K input + 3K output tokens

Llama 4 Scout$0.0059Fits in one request
Qwen3.5-Flash$0.0040Fits in one request

Qwen3.5-Flash costs less

Agent loop

200K input + 10K output tokens

Llama 4 Scout$0.023Fits in one request
Qwen3.5-Flash$0.016Fits in one request

Qwen3.5-Flash costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of Llama 4 Scout and Qwen3.5-Flash
SpecificationMeta Llama 4 ScoutQwen3.5-Flash
MakerMeta LlamaQwen
WeightsOpen weightsNo public weights listed
ReleasedApril 5, 2025February 25, 2026
Context window1.31M1M
Input $/Mtok$0.10$0.065
Output $/Mtok$0.30$0.26
Coding index9.16.2
Independent measurements behind it11
Cost per solved task$1.276$0.40
SourceOpenRouter listingOpenRouter listing

Questions

Is Llama 4 Scout better than Qwen3.5-Flash for coding?

Pick Llama 4 Scout for coding: it scores 9.1 on the Vibe Coding Index against 6.2 for Qwen3.5-Flash. Pick Qwen3.5-Flash when output tokens cost 1.2x less, or when a solved task costs $0.40 against $1.276.

Which is cheaper, Llama 4 Scout or Qwen3.5-Flash?

Llama 4 Scout costs $0.10 per million input tokens and $0.30 per million output tokens; Qwen3.5-Flash costs $0.065 and $0.26. Per solved SWE-bench Verified task, Llama 4 Scout costs $1.276 and Qwen3.5-Flash $0.40.

Which has the bigger context window, Llama 4 Scout or Qwen3.5-Flash?

Llama 4 Scout, with 1.31M tokens against 1M.

How many benchmarks is the Llama 4 Scout vs Qwen3.5-Flash comparison based on?

45 benchmarks have a result for both models as of October 21, 2026. Llama 4 Scout's coding index rests on 1 independent benchmark and Qwen3.5-Flash's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next