Noometry

Model comparison

Nova 2 Lite vs Qwen3.8 27B

Qwen3.8 27B is the stronger model overall, scoring 46.0 to 39.7 on the Noometry Index.

Last verified . 17 shared benchmarks.

Nova 2 Lite Amazon

39.7

Rank #161 Confirmed

Qwen3.8 27B Alibaba (Qwen)

46.0

Rank #68 Confirmed

Summary

  • They share 17 benchmarks with published results for both. Nova 2 Lite scores higher in 2 categories and Qwen3.8 27B in 7 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Qwen3.8 27B leads 41.0 to 27.5.
  • Nova 2 Lite is cheaper at $0.30 / $2.50 per million input/output tokens, against $0.99 / $1.49 for Qwen3.8 27B.
  • Nova 2 Lite accepts more context: 1M tokens versus 262K.
  • Qwen3.8 27B has downloadable open weights; the other is API-only.

Side by side

Nova 2 LiteQwen3.8 27B
ProviderAmazonAlibaba (Qwen)
Noometry Index39.746.0
Released2025-12-012026-08-14
WeightsProprietaryOpen
Context window1M262K
Max output64K33K
Input $ / M tokens$0.30$0.99
Output $ / M tokens$2.50$1.49
Results tracked1931

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Qwen3.8 27B leads

Nova 2 Lite: 40.7 (#134), Qwen3.8 27B: 50.5 (#44)

BenchmarkNova 2 LiteQwen3.8 27B
LMArena Coding13851482
LMArena WebDev—1593
SciCode—46.6%

Agentic & Tool Use Qwen3.8 27B leads

Nova 2 Lite: 24.1 (#121), Qwen3.8 27B: 32.9 (#57)

BenchmarkNova 2 LiteQwen3.8 27B
APEX-Agents—47.5%
Berkeley Function Calling Leaderboard27.1%—

Reasoning Qwen3.8 27B leads

Nova 2 Lite: 27.5 (#118), Qwen3.8 27B: 41.0 (#54)

BenchmarkNova 2 LiteQwen3.8 27B
LMArena Hard Prompts13641460
ARC-AGI-2—42.4%
NYT Connections (extended)—54.5%
ARC-AGI-1—87.5%
CritPt—5.4%
DTBench—88%
LMCA—41.4%
Surface Evolver Bench—45%
Epoch Capabilities Index—149.38

Math Too close to call

Nova 2 Lite: 37.5 (#156), Qwen3.8 27B: 37.1 (#161)

BenchmarkNova 2 LiteQwen3.8 27B
LMArena Math13591456
ProofBench—16%

Knowledge Nova 2 Lite leads

Nova 2 Lite: 43.0 (#94), Qwen3.8 27B: 41.6 (#109)

BenchmarkNova 2 LiteQwen3.8 27B
LMArena Expert13581482
Vectara Hallucination Rate5.1%—

Multimodal Not comparable

Nova 2 Lite: —, Qwen3.8 27B: 41.3 (#37)

BenchmarkNova 2 LiteQwen3.8 27B
LMArena Vision—1271

Multilingual Qwen3.8 27B leads

Nova 2 Lite: 47.1 (#153), Qwen3.8 27B: 53.7 (#60)

BenchmarkNova 2 LiteQwen3.8 27B
LMArena Non-English13371430
LMArena Chinese13641504
LMArena French13811465
LMArena German13431438
LMArena Japanese12711384
LMArena Korean12841393
LMArena Russian13431415
LMArena Spanish13731448

Instruction Following Qwen3.8 27B leads

Nova 2 Lite: 70.5 (#161), Qwen3.8 27B: 75.8 (#53)

BenchmarkNova 2 LiteQwen3.8 27B
LMArena Instruction Following13351439

Long Context Qwen3.8 27B leads

Nova 2 Lite: 40.6 (#150), Qwen3.8 27B: 44.3 (#70)

BenchmarkNova 2 LiteQwen3.8 27B
LMArena Longer Query13351450

Writing & Preference Qwen3.8 27B leads

Nova 2 Lite: 53.9 (#154), Qwen3.8 27B: 65.8 (#43)

BenchmarkNova 2 LiteQwen3.8 27B
LMArena Text13621441
LMArena Creative Writing12911384
LMArena Multi-Turn13381441
EQ-Bench Creative Writing—1671

Frequently asked questions

Is Nova 2 Lite better than Qwen3.8 27B?

Qwen3.8 27B is the stronger model overall, scoring 46.0 to 39.7 on the Noometry Index.

Which is cheaper, Nova 2 Lite or Qwen3.8 27B?

Nova 2 Lite is cheaper. It lists at $0.30 per million input tokens and $2.50 per million output tokens; Qwen3.8 27B lists at $0.99 and $1.49.

Is Nova 2 Lite or Qwen3.8 27B better for coding?

Qwen3.8 27B scores higher on coding benchmarks: 50.5 versus 40.7 in the Noometry coding category.

Which has the bigger context window?

Nova 2 Lite does, with 1M tokens against 262K.

How many benchmarks do Nova 2 Lite and Qwen3.8 27B share?

17 benchmarks have published results for both models. Nova 2 Lite has 19 scored results on Noometry and Qwen3.8 27B has 31.

Related comparisons

Go deeper