Enter opens the highlighted result. Esc closes.

Nova Lite 1.0 vs Qwen2.5 72B Instruct

Qwen2.5 72B Instruct leads Nova Lite 1.0 on the Artificial Analysis Intelligence Index, 7.7 to 6.7, and Nova Lite 1.0 costs 1.7x less per output token ($0.24 vs $0.40 per million). Updated October 21, 2026; 38 benchmarks have a result for both.

Replace Nova Lite 1.0 Replace Qwen2.5 72B Instruct
Nova Lite 1.0Amazon
Coding
No data
Overall
No data
Output $/M
$0.24
1
Qwen2.5 72B InstructQwen
Coding
No data
Overall
26.5
Output $/M
$0.40
37
vs boards won of 38 shared

Nova Lite 1.0 or Qwen2.5 72B Instruct: which to pick

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

Pick Nova Lite 1.0 for output tokens 1.7x cheaper ($0.24 vs $0.40 per million) and a larger context window (300K against 33K tokens).

Model ANova Lite 1.0AmazonNot benchmarkedNo coding benchmarkNo cost per solved taskModel BQwen2.5 72B InstructQwenNot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
38
Nova Lite 1.0 only
3
Qwen2.5 72B Instruct only
15
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
Nova Lite 1.0 not benchmarked, Qwen2.5 72B Instruct not benchmarked.
Widest category
Agentic work, 17.0 points apart on average over 1 shared benchmark.
Widest single gap
τ²-Bench Telecom: Nova Lite 1.0 17.5% against Qwen2.5 72B Instruct 34.5%, 17.0 points apart on a 0 to 100 scale.
Output price
Nova Lite 1.0 $0.24 against Qwen2.5 72B Instruct $0.40 per million output tokens, 1.7x apart.
Context window
Nova Lite 1.0 takes 300K 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
CategoryNova Lite 1.0BarsQwen2.5 72B InstructTies
Coding04
Intelligence01
Agentic work01
Reasoning14
Math04
Knowledge06
General017
All shared boards137
CodingLike-for-likeSix of these boards feed the Vibe Coding Index04
Coding benchmarks for Nova Lite 1.0 and Qwen2.5 72B Instruct
BenchmarkNova Lite 1.0Qwen2.5 72B InstructLeader
LiveCodeBench (Artificial Analysis run)16.7%27.6%Qwen2.5 72B Instruct
Terminal-Bench Hard (Artificial Analysis run)0.8%4.5%Qwen2.5 72B Instruct
WeirdMLNot measured16.0%Only Qwen2.5 72B Instruct measured
LMArena Text: coding12391292Qwen2.5 72B Instruct
LMArena Text: software and IT12411292Qwen2.5 72B Instruct
Neither model has a result on the other 34 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for Nova Lite 1.0 and Qwen2.5 72B Instruct
BenchmarkNova Lite 1.0Qwen2.5 72B InstructLeader
AA Intelligence Index6.77.7Qwen2.5 72B Instruct
Epoch Capabilities IndexNot measured129.0Only Qwen2.5 72B Instruct measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for Nova Lite 1.0 and Qwen2.5 72B Instruct
BenchmarkNova Lite 1.0Qwen2.5 72B InstructLeader
τ²-Bench Telecom17.5%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
Neither model has a result on the other 37 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index14
Reasoning benchmarks for Nova Lite 1.0 and Qwen2.5 72B Instruct
BenchmarkNova Lite 1.0Qwen2.5 72B InstructLeader
GPQA Diamond (Artificial Analysis run)43.3%49.1%Qwen2.5 72B Instruct
Humanity's Last Exam (Artificial Analysis run)4.3%3.6%Nova Lite 1.0
GPQA Diamond (Epoch AI run)Not measured49.1%Only Qwen2.5 72B Instruct measured
BIG-Bench HardNot measured79.8%Only Qwen2.5 72B Instruct measured
LMArena Text: hard prompts12201271Qwen2.5 72B Instruct
LMArena Text: expert12011245Qwen2.5 72B Instruct
LMArena Text: hard prompts in English12261279Qwen2.5 72B Instruct
Neither model has a result on the other 15 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index04
Math benchmarks for Nova Lite 1.0 and Qwen2.5 72B Instruct
BenchmarkNova Lite 1.0Qwen2.5 72B InstructLeader
AA Math Index7.014.0Qwen2.5 72B Instruct
AIME 2025 (Artificial Analysis run)7.0%14.0%Qwen2.5 72B Instruct
OTIS Mock AIMENot measured8.1%Only Qwen2.5 72B Instruct measured
MATH level 5Not measured63.2%Only Qwen2.5 72B Instruct measured
LMArena Text: math12271283Qwen2.5 72B Instruct
LMArena Text: mathematical work12351290Qwen2.5 72B Instruct
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index06
Knowledge benchmarks for Nova Lite 1.0 and Qwen2.5 72B Instruct
BenchmarkNova Lite 1.0Qwen2.5 72B InstructLeader
MMLU Pro (Artificial Analysis run)59.0%72.0%Qwen2.5 72B Instruct
MMLU77.0%85.3%Qwen2.5 72B Instruct
AI2 Reasoning ChallengeNot measured94.5%Only Qwen2.5 72B Instruct measured
LMArena Text: science12321275Qwen2.5 72B Instruct
LMArena Text: medicine and healthcare12131258Qwen2.5 72B Instruct
LMArena Text: legal and government12381288Qwen2.5 72B Instruct
LMArena Text: business and finance12191253Qwen2.5 72B Instruct
Neither model has a result on the other 11 boards here
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Web design benchmarks for Nova Lite 1.0 and Qwen2.5 72B Instruct
BenchmarkNova Lite 1.0Qwen2.5 72B InstructLeader
Neither model has a result on 9 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index017
General benchmarks for Nova Lite 1.0 and Qwen2.5 72B Instruct
BenchmarkNova Lite 1.0Qwen2.5 72B InstructLeader
IFBench34.1%36.9%Qwen2.5 72B Instruct
Long context reasoning (AA-LCR)19.0%Not measuredOnly Nova Lite 1.0 measured
DTBenchNot measured62.9%Only Qwen2.5 72B Instruct measured
LMCANot measured13.4%Only Qwen2.5 72B Instruct 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 Text12291269Qwen2.5 72B Instruct
LMArena Text: instruction following12051254Qwen2.5 72B Instruct
LMArena Text: creative writing11971221Qwen2.5 72B Instruct
LMArena Vision990Not measuredOnly Nova Lite 1.0 measured
LMArena Text: English12431283Qwen2.5 72B Instruct
LMArena Text: non-English12081252Qwen2.5 72B Instruct
LMArena Text: Chinese12251273Qwen2.5 72B Instruct
LMArena Text: German12291234Qwen2.5 72B Instruct
LMArena Text: Japanese11531180Qwen2.5 72B Instruct
LMArena Text: Korean11531188Qwen2.5 72B Instruct
LMArena Text: French12381280Qwen2.5 72B Instruct
LMArena Text: Spanish12261255Qwen2.5 72B Instruct
LMArena Text: Russian12161264Qwen2.5 72B Instruct
LMArena Text: multi-turn11991272Qwen2.5 72B Instruct
LMArena Text: longer queries12341281Qwen2.5 72B Instruct
LMArena Text: writing and language12111247Qwen2.5 72B Instruct
LMArena Text: entertainment and media11761211Qwen2.5 72B Instruct
LMArena Vision: English999Not measuredOnly Nova Lite 1.0 measured
Neither model has a result on the other 21 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Nova Lite 1.0 and Qwen2.5 72B Instruct
BenchmarkNova Lite 1.0Qwen2.5 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

Nova Lite 1.0$0.0002Fits in one request
Qwen2.5 72B Instruct$0.0006Fits in one request

Nova Lite 1.0 costs less

Repository review

50K input + 3K output tokens

Nova Lite 1.0$0.0037Fits in one request
Qwen2.5 72B Instruct$0.019Exceeds the context window

Nova Lite 1.0 costs less

Agent loop

200K input + 10K output tokens

Nova Lite 1.0$0.014Fits in one request
Qwen2.5 72B Instruct$0.076Exceeds the context window

Nova Lite 1.0 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 Nova Lite 1.0 and Qwen2.5 72B Instruct
SpecificationAmazon Nova Lite 1.0Qwen2.5 72B Instruct
MakerAmazonQwen
WeightsNo public weights listedOpen weights
ReleasedDecember 5, 2024September 19, 2024
Context window300K33K
Input $/Mtok$0.06$0.36
Output $/Mtok$0.24$0.40
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it00
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Nova Lite 1.0 better than Qwen2.5 72B Instruct for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Nova Lite 1.0 for output tokens 1.7x cheaper ($0.24 vs $0.40 per million) and a larger context window (300K against 33K tokens).

Which is cheaper, Nova Lite 1.0 or Qwen2.5 72B Instruct?

Nova Lite 1.0 costs $0.06 per million input tokens and $0.24 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, Nova Lite 1.0 or Qwen2.5 72B Instruct?

Nova Lite 1.0, with 300K tokens against 33K.

How many benchmarks is the Nova Lite 1.0 vs Qwen2.5 72B Instruct comparison based on?

38 benchmarks have a result for both models as of October 21, 2026. Nova Lite 1.0'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