LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

GPT-4 vs Qwen3.5-Flash

Updated October 21, 2026. 31 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace GPT-4 Replace Qwen3.5-Flash
GPT-4OpenAI
Coding
22.4
Overall
6.7
Output $/M
$60.00
0
Qwen3.5-FlashQwen
Coding
6.2
Overall
No data
Output $/M
$0.26
31
vs boards won of 31 shared

Decision reading

Pick GPT-4 for coding: it scores 22.4 on the Vibe Coding Index against 6.2 for Qwen3.5-Flash.

Pick Qwen3.5-Flash when output tokens cost 230.8x less, or when the context window is larger, 1M against 8K tokens.

Model AGPT-4OpenAI22.4/1001 independent benchmarkNo cost per solved taskModel BQwen3.5-FlashQwen6.2/1001 independent benchmark$0.40 per solved task
Shared results
31
GPT-4 only
14
Qwen3.5-Flash only
24
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
GPT-4 22.4 from 1 benchmark, Qwen3.5-Flash 6.2 from 1 benchmark.
Widest category
Knowledge, 39.8 points apart on average over 4 shared benchmarks.
Widest single gap
LMArena Text: medicine and healthcare: GPT-4 1145 against Qwen3.5-Flash 1407, 43.6 points apart on a 0 to 100 scale.
Output price
GPT-4 $60.00 against Qwen3.5-Flash $0.26 per million output tokens, 230.8x apart.
Context window
GPT-4 takes 8K 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
CategoryGPT-4BarsQwen3.5-FlashTies
Coding02
Intelligence01
Reasoning03
Math02
Knowledge04
General019
All shared boards031
CodingLike-for-likeFour of these boards feed the Vibe Coding Index02
Coding benchmarks for GPT-4 and Qwen3.5-Flash
BenchmarkGPT-4Qwen3.5-FlashLeader
SWE-bench Verified22.4%Not measuredOnly GPT-4 measured
SWE-bench Lite39.7%Not measuredOnly GPT-4 measured
SWE-bench (full test set)12.5%Not measuredOnly GPT-4 measured
LMArena WebDevNot measured1244Only Qwen3.5-Flash measured
AA Coding Index13.1Not measuredOnly GPT-4 measured
LiveCodeBench (Vals AI)Not measured83.3%Only Qwen3.5-Flash measured
SWE-bench Verified (Vals AI)Not measured64.4%Only Qwen3.5-Flash measured
ALE-BenchNot measured265.9Only Qwen3.5-Flash measured
LMArena Text: coding12091412Qwen3.5-Flash
LMArena Text: software and IT11951416Qwen3.5-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 26 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for GPT-4 and Qwen3.5-Flash
BenchmarkGPT-4Qwen3.5-FlashLeader
AA Intelligence Index6.7Not measuredOnly GPT-4 measured
Epoch Capabilities Index125.7143.5Qwen3.5-Flash
Agentic workNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Agentic work benchmarks for GPT-4 and Qwen3.5-Flash
BenchmarkGPT-4Qwen3.5-FlashLeader
Neither model has a result on 39 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index03
Reasoning benchmarks for GPT-4 and Qwen3.5-Flash
BenchmarkGPT-4Qwen3.5-FlashLeader
GPQA DiamondNot measured82.8%Only Qwen3.5-Flash measured
MMMU Pro (Vals AI)Not measured81.9%Only Qwen3.5-Flash measured
GPQA Diamond (Artificial Analysis run)34.9%Not measuredOnly GPT-4 measured
GPQA Diamond (Epoch AI run)Not measured82.3%Only Qwen3.5-Flash measured
Chess puzzlesNot measured14.0%Only Qwen3.5-Flash measured
Mystery game puzzlesNot measured20.0%Only Qwen3.5-Flash measured
BIG-Bench Hard75.1%Not measuredOnly GPT-4 measured
LMArena Text: hard prompts12001403Qwen3.5-Flash
LMArena Text: expert11491407Qwen3.5-Flash
LMArena Text: hard prompts in English12091409Qwen3.5-Flash
Neither model has a result on the other 11 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index02
Math benchmarks for GPT-4 and Qwen3.5-Flash
BenchmarkGPT-4Qwen3.5-FlashLeader
AIMENot measured92.5%Only Qwen3.5-Flash measured
OTIS Mock AIMENot measured84.4%Only Qwen3.5-Flash measured
FrontierMath, tiers 1 to 3Not measured18.2%Only Qwen3.5-Flash measured
GSM8K92.0%Not measuredOnly GPT-4 measured
LMArena Text: math12301407Qwen3.5-Flash
LMArena Text: mathematical work12281413Qwen3.5-Flash
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
Knowledge benchmarks for GPT-4 and Qwen3.5-Flash
BenchmarkGPT-4Qwen3.5-FlashLeader
MMLU ProNot measured84.1%Only Qwen3.5-Flash measured
LegalBench (Vals AI)Not measured84.3%Only Qwen3.5-Flash measured
CaseLaw (Vals AI)Not measured56.0%Only Qwen3.5-Flash measured
CorpFin (Vals AI)Not measured63.6%Only Qwen3.5-Flash measured
TaxEval (Vals AI)Not measured72.2%Only Qwen3.5-Flash measured
MortgageTax (Vals AI)Not measured67.4%Only Qwen3.5-Flash measured
MedCode (Vals AI)Not measured33.0%Only Qwen3.5-Flash measured
MMLU Pro (Artificial Analysis run)56.2%Not measuredOnly GPT-4 measured
SimpleQA VerifiedNot measured20.3%Only Qwen3.5-Flash measured
MMLU86.4%Not measuredOnly GPT-4 measured
AI2 Reasoning Challenge96.3%Not measuredOnly GPT-4 measured
LMArena Text: science11771415Qwen3.5-Flash
LMArena Text: medicine and healthcare11451407Qwen3.5-Flash
LMArena Text: legal and government11911398Qwen3.5-Flash
LMArena Text: business and finance11451394Qwen3.5-Flash
Neither model has a result on the other 2 boards here
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Web design benchmarks for GPT-4 and Qwen3.5-Flash
BenchmarkGPT-4Qwen3.5-FlashLeader
Neither model has a result on 9 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index019
General benchmarks for GPT-4 and Qwen3.5-Flash
BenchmarkGPT-4Qwen3.5-FlashLeader
SAGE (Vals AI)Not measured42.5%Only Qwen3.5-Flash measured
MedScribe (Vals AI)Not measured70.6%Only Qwen3.5-Flash measured
IFBench33.2%Not measuredOnly GPT-4 measured
DTBench62.7%82.9%Qwen3.5-Flash
LMCA17.1%29.1%Qwen3.5-Flash
HellaSwag95.3%Not measuredOnly GPT-4 measured
WinoGrande87.5%Not measuredOnly GPT-4 measured
LMArena Text12061397Qwen3.5-Flash
LMArena Text: instruction following12011374Qwen3.5-Flash
LMArena Text: creative writing11911343Qwen3.5-Flash
LMArena Text: English12201404Qwen3.5-Flash
LMArena Text: non-English11941385Qwen3.5-Flash
LMArena Text: Chinese11841445Qwen3.5-Flash
LMArena Text: German11981390Qwen3.5-Flash
LMArena Text: Japanese11371369Qwen3.5-Flash
LMArena Text: Korean10871344Qwen3.5-Flash
LMArena Text: French12201413Qwen3.5-Flash
LMArena Text: Spanish11971400Qwen3.5-Flash
LMArena Text: Russian11951380Qwen3.5-Flash
LMArena Text: Polish12651376Qwen3.5-Flash
LMArena Text: multi-turn12061393Qwen3.5-Flash
LMArena Text: longer queries11901392Qwen3.5-Flash
LMArena Text: writing and language12191369Qwen3.5-Flash
LMArena Text: entertainment and media11841347Qwen3.5-Flash
Neither model has a result on the other 22 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-4 and Qwen3.5-Flash
BenchmarkGPT-4Qwen3.5-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

GPT-4$0.060Fits in one request
Qwen3.5-Flash$0.0002Fits in one request

Qwen3.5-Flash costs less

Repository review

50K input + 3K output tokens

GPT-4$1.68Exceeds the context window
Qwen3.5-Flash$0.0040Fits in one request

Qwen3.5-Flash costs less

Agent loop

200K input + 10K output tokens

GPT-4$6.60Exceeds the context window
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 GPT-4 and Qwen3.5-Flash
SpecificationOpenAI GPT-4Qwen3.5-Flash
MakerOpenAIQwen
WeightsNo public weights listedNo public weights listed
ReleasedMay 28, 2023February 25, 2026
Context window8K1M
Input $/Mtok$30.00$0.065
Output $/Mtok$60.00$0.26
Coding index22.46.2
Independent measurements behind it11
Cost per solved taskNot computable$0.40
SourceOpenRouter listingOpenRouter listing

Questions

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

Pick GPT-4 for coding: it scores 22.4 on the Vibe Coding Index against 6.2 for Qwen3.5-Flash. Pick Qwen3.5-Flash when output tokens cost 230.8x less, or when the context window is larger, 1M against 8K tokens.

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

GPT-4 costs $30.00 per million input tokens and $60.00 per million output tokens; Qwen3.5-Flash costs $0.065 and $0.26. Cost per solved task is known for Qwen3.5-Flash only.

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

Qwen3.5-Flash, with 1M tokens against 8K.

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

31 benchmarks have a result for both models as of October 21, 2026. GPT-4'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