Enter opens the highlighted result. Esc closes.

GPT-4.1 Nano vs Qwen2.5 72B Instruct

GPT-4.1 Nano leads Qwen2.5 72B Instruct on the Artificial Analysis Intelligence Index, 7.8 to 7.7, at about the same output price ($0.40 vs $0.40 per million). Updated October 22, 2026; 41 benchmarks have a result for both.

Replace GPT-4.1 Nano Replace Qwen2.5 72B Instruct
GPT-4.1 NanoOpenAI
Coding
No data
Overall
24.7
Output $/M
$0.40
30
Qwen2.5 72B InstructQwen
Coding
No data
Overall
26.5
Output $/M
$0.40
11
vs boards won of 41 shared

GPT-4.1 Nano or Qwen2.5 72B Instruct: which to pick

Neither model has a coding score yet, so the two cannot be ranked on coding.

Pick GPT-4.1 Nano for a larger context window (1.05M against 33K tokens) and work measured by Terminal-Bench Hard (Artificial Analysis run) or τ²-Bench Telecom, where it leads.

Model AGPT-4.1 NanoOpenAINot benchmarkedNo coding benchmarkNo cost per solved taskModel BQwen2.5 72B InstructQwenNot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
41
GPT-4.1 Nano only
30
Qwen2.5 72B Instruct only
12
Like-for-like categories
7 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

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-4.1 Nano not benchmarked, Qwen2.5 72B Instruct not benchmarked.
Widest category
Agentic work, 17.3 points apart on average over 1 shared benchmark.
Widest single gap
OTIS Mock AIME: GPT-4.1 Nano 28.9% against Qwen2.5 72B Instruct 8.1%, 20.8 points apart on a 0 to 100 scale.
Output price
Both cost $0.40 per million output tokens.
Context window
GPT-4.1 Nano takes 1.05M tokens, Qwen2.5 72B Instruct 33K.

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-4.1 NanoBarsQwen2.5 72B InstructTies
Coding41
Intelligence20
Agentic work01
Reasoning51
Math42
Knowledge41
General115
All shared boards3011
CodingLike-for-likeSix of these boards feed the Vibe Coding Index41
Coding benchmarks for GPT-4.1 Nano and Qwen2.5 72B Instruct
BenchmarkGPT-4.1 NanoQwen2.5 72B InstructLeader
Aider polyglot8.9%Not measuredOnly GPT-4.1 Nano measured
AA Coding Index11.1Not measuredOnly GPT-4.1 Nano measured
LiveCodeBench (Vals AI)42.7%Not measuredOnly GPT-4.1 Nano measured
LiveCodeBench (Artificial Analysis run)32.6%27.6%GPT-4.1 Nano
Terminal-Bench Hard (Artificial Analysis run)3.8%4.5%Qwen2.5 72B Instruct
WeirdML19.0%16.0%GPT-4.1 Nano
LMArena Text: coding13061292GPT-4.1 Nano
LMArena Text: software and IT13051292GPT-4.1 Nano
Neither model has a result on the other 31 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for GPT-4.1 Nano and Qwen2.5 72B Instruct
BenchmarkGPT-4.1 NanoQwen2.5 72B InstructLeader
AA Intelligence Index7.87.7GPT-4.1 Nano
Epoch Capabilities Index129.5129.0GPT-4.1 Nano
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for GPT-4.1 Nano and Qwen2.5 72B Instruct
BenchmarkGPT-4.1 NanoQwen2.5 72B InstructLeader
τ²-Bench Telecom17.3%34.5%Qwen2.5 72B Instruct
METR time horizonsNot measured35.8%Only Qwen2.5 72B Instruct measured
BALROGNot measured16.2%Only Qwen2.5 72B Instruct measured
TheAgentCompanyNot measured5.7%Only Qwen2.5 72B Instruct measured
Berkeley Function Calling Leaderboard (BFCL)33.1%Not measuredOnly GPT-4.1 Nano measured
Neither model has a result on the other 36 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index51
Reasoning benchmarks for GPT-4.1 Nano and Qwen2.5 72B Instruct
BenchmarkGPT-4.1 NanoQwen2.5 72B InstructLeader
GPQA Diamond50.8%Not measuredOnly GPT-4.1 Nano measured
MMMU Pro (Vals AI)55.0%Not measuredOnly GPT-4.1 Nano measured
GPQA Diamond (Artificial Analysis run)51.2%49.1%GPT-4.1 Nano
Humanity's Last Exam (Artificial Analysis run)3.8%3.6%GPT-4.1 Nano
GPQA Diamond (Epoch AI run)48.9%49.1%Qwen2.5 72B Instruct
ARC-AGI0.0%Not measuredOnly GPT-4.1 Nano measured
ARC-AGI-20.0%Not measuredOnly GPT-4.1 Nano measured
Chess puzzles0.0%Not measuredOnly GPT-4.1 Nano measured
BIG-Bench HardNot measured79.8%Only Qwen2.5 72B Instruct measured
LMArena Text: hard prompts12861271GPT-4.1 Nano
LMArena Text: expert12711245GPT-4.1 Nano
LMArena Text: hard prompts in English12931279GPT-4.1 Nano
Kagi LLM Benchmark33.3%Not measuredOnly GPT-4.1 Nano measured
Neither model has a result on the other 9 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index42
Math benchmarks for GPT-4.1 Nano and Qwen2.5 72B Instruct
BenchmarkGPT-4.1 NanoQwen2.5 72B InstructLeader
AIME26.5%Not measuredOnly GPT-4.1 Nano measured
MATH 500 (Vals AI)80.2%Not measuredOnly GPT-4.1 Nano measured
MGSM (Vals AI)69.3%Not measuredOnly GPT-4.1 Nano measured
AA Math Index24.014.0GPT-4.1 Nano
AIME 2025 (Artificial Analysis run)24.0%14.0%GPT-4.1 Nano
OTIS Mock AIME28.9%8.1%GPT-4.1 Nano
MATH level 570.0%63.2%GPT-4.1 Nano
LMArena Text: math12741283Qwen2.5 72B Instruct
LMArena Text: mathematical work12741290Qwen2.5 72B Instruct
Neither model has a result on the other 4 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index41
Knowledge benchmarks for GPT-4.1 Nano and Qwen2.5 72B Instruct
BenchmarkGPT-4.1 NanoQwen2.5 72B InstructLeader
MMLU Pro63.5%Not measuredOnly GPT-4.1 Nano measured
LegalBench (Vals AI)61.1%Not measuredOnly GPT-4.1 Nano measured
CorpFin (Vals AI)42.1%Not measuredOnly GPT-4.1 Nano measured
TaxEval (Vals AI)60.8%Not measuredOnly GPT-4.1 Nano measured
MortgageTax (Vals AI)52.8%Not measuredOnly GPT-4.1 Nano measured
MedQA (Vals AI)68.2%Not measuredOnly GPT-4.1 Nano measured
MMLU Pro (Artificial Analysis run)65.7%72.0%Qwen2.5 72B Instruct
SimpleQA Verified6.0%Not measuredOnly GPT-4.1 Nano measured
MMLUNot measured85.3%Only Qwen2.5 72B Instruct measured
AI2 Reasoning ChallengeNot measured94.5%Only Qwen2.5 72B Instruct measured
LMArena Text: science12971275GPT-4.1 Nano
LMArena Text: medicine and healthcare12761258GPT-4.1 Nano
LMArena Text: legal and government13001288GPT-4.1 Nano
LMArena Text: business and finance12671253GPT-4.1 Nano
Neither model has a result on the other 4 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
Web design benchmarks for GPT-4.1 Nano and Qwen2.5 72B Instruct
BenchmarkGPT-4.1 NanoQwen2.5 72B InstructLeader
Design Arena: websites973Not measuredOnly GPT-4.1 Nano measured
Design Arena: UI components913Not measuredOnly GPT-4.1 Nano measured
Design Arena: game dev970Not measuredOnly GPT-4.1 Nano measured
Design Arena: data viz891Not measuredOnly GPT-4.1 Nano measured
Design Arena: 3D924Not measuredOnly GPT-4.1 Nano measured
Design Arena: all code categories962Not measuredOnly GPT-4.1 Nano measured
Neither model has a result on the other 3 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index115
General benchmarks for GPT-4.1 Nano and Qwen2.5 72B Instruct
BenchmarkGPT-4.1 NanoQwen2.5 72B InstructLeader
IFBench32.0%36.9%Qwen2.5 72B Instruct
Long context reasoning (AA-LCR)20.3%Not measuredOnly GPT-4.1 Nano measured
DTBench52.5%62.9%Qwen2.5 72B Instruct
LMCA5.5%13.4%Qwen2.5 72B Instruct
Fiction.LiveBench25.0%Not measuredOnly GPT-4.1 Nano measured
HellaSwagNot measured84.8%Only Qwen2.5 72B Instruct measured
WinoGrandeNot measured82.3%Only Qwen2.5 72B Instruct measured
PIQANot measured82.6%Only Qwen2.5 72B Instruct measured
LMArena Text12851269GPT-4.1 Nano
LMArena Text: instruction following12661254GPT-4.1 Nano
LMArena Text: creative writing12601221GPT-4.1 Nano
LMArena Vision1063Not measuredOnly GPT-4.1 Nano measured
LMArena Text: English13031283GPT-4.1 Nano
LMArena Text: non-English12601252GPT-4.1 Nano
LMArena Text: Chinese12701273Qwen2.5 72B Instruct
LMArena Text: German12881234GPT-4.1 Nano
LMArena Text: Japanese11981180GPT-4.1 Nano
LMArena Text: KoreanNot measured1188Only Qwen2.5 72B Instruct measured
LMArena Text: FrenchNot measured1280Only Qwen2.5 72B Instruct measured
LMArena Text: SpanishNot measured1255Only Qwen2.5 72B Instruct measured
LMArena Text: Russian12611264Qwen2.5 72B Instruct
LMArena Text: multi-turn12771272GPT-4.1 Nano
LMArena Text: longer queries12831281GPT-4.1 Nano
LMArena Text: writing and language12611247GPT-4.1 Nano
LMArena Text: entertainment and media12431211GPT-4.1 Nano
LMArena Vision: English1101Not measuredOnly GPT-4.1 Nano measured
Neither model has a result on the other 20 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-4.1 Nano and Qwen2.5 72B Instruct
BenchmarkGPT-4.1 NanoQwen2.5 72B InstructLeader
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-4.1 Nano$0.0003Fits in one request
Qwen2.5 72B Instruct$0.0006Fits in one request

GPT-4.1 Nano costs less

Repository review

50K input + 3K output tokens

GPT-4.1 Nano$0.0062Fits in one request
Qwen2.5 72B Instruct$0.019Exceeds the context window

GPT-4.1 Nano costs less

Agent loop

200K input + 10K output tokens

GPT-4.1 Nano$0.024Fits in one request
Qwen2.5 72B Instruct$0.076Exceeds the context window

GPT-4.1 Nano 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-4.1 Nano and Qwen2.5 72B Instruct
SpecificationOpenAI GPT-4.1 NanoQwen2.5 72B Instruct
MakerOpenAIQwen
WeightsNo public weights listedOpen weights
ReleasedApril 14, 2025September 19, 2024
Context window1.05M33K
Input $/Mtok$0.10$0.36
Output $/Mtok$0.40$0.40
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it00
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-4.1 Nano better than Qwen2.5 72B Instruct for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Pick GPT-4.1 Nano for a larger context window (1.05M against 33K tokens) and work measured by Terminal-Bench Hard (Artificial Analysis run) or τ²-Bench Telecom, where it leads.

Which is cheaper, GPT-4.1 Nano or Qwen2.5 72B Instruct?

GPT-4.1 Nano costs $0.10 per million input tokens and $0.40 per million output tokens; Qwen2.5 72B Instruct costs $0.36 and $0.40. Cost per solved task is known for neither of them.

Which has the bigger context window, GPT-4.1 Nano or Qwen2.5 72B Instruct?

GPT-4.1 Nano, with 1.05M tokens against 33K.

How many benchmarks is the GPT-4.1 Nano vs Qwen2.5 72B Instruct comparison based on?

41 benchmarks have a result for both models as of October 22, 2026. GPT-4.1 Nano's coding index rests on 0 independent benchmarks and Qwen2.5 72B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.

Compare next