Noometry

Model comparison

Qwen3.8 27B vs Qwen3 8B

Qwen3.8 27B is the stronger model overall, scoring 46.0 to 33.7 on the Noometry Index. Qwen3 8B costs 2.0× less per token, which makes it the better buy when Qwen3.8 27B's lead doesn't matter for your workload.

Last verified . 5 shared benchmarks.

Qwen3.8 27B Alibaba (Qwen)

46.0

Rank #68 Confirmed

Qwen3 8B Alibaba (Qwen)

33.7

Rank #238 Confirmed

Summary

  • They share 5 benchmarks with published results for both. Qwen3.8 27B scores higher in 6 categories and Qwen3 8B in 0 categories; 6 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Qwen3.8 27B leads 41.0 to 16.6.
  • The biggest single-benchmark swing is LMCA: 41.4% for Qwen3.8 27B and 8.8% for Qwen3 8B.
  • Qwen3 8B is cheaper at $0.18 / $0.70 per million input/output tokens, against $0.04 / $2.30 for Qwen3.8 27B.
  • Qwen3.8 27B accepts more context: 262K tokens versus 131K.

Side by side

Qwen3.8 27BQwen3 8B
ProviderAlibaba (Qwen)Alibaba (Qwen)
Noometry Index46.033.7
Released2026-08-142025-04
WeightsOpenOpen
Context window262K131K
Max output33K8K
Input $ / M tokens$0.04$0.18
Output $ / M tokens$2.30$0.70
Results tracked3111

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

Category by category

Coding Qwen3.8 27B leads

Qwen3.8 27B: 50.5 (#44), Qwen3 8B: 34.0 (#248)

BenchmarkQwen3.8 27BQwen3 8B
SciCode46.6%22.6%
LMArena WebDev1593—
LMArena Coding1482—

Agentic & Tool Use Qwen3.8 27B leads

Qwen3.8 27B: 32.9 (#57), Qwen3 8B: 30.2 (#78)

BenchmarkQwen3.8 27BQwen3 8B
APEX-Agents47.5%—
Berkeley Function Calling Leaderboard—42.6%

Reasoning Qwen3.8 27B leads

Qwen3.8 27B: 41.0 (#54), Qwen3 8B: 16.6 (#303)

BenchmarkQwen3.8 27BQwen3 8B
CritPt5.4%0%
DTBench88%59.7%
LMCA41.4%8.8%
Epoch Capabilities Index149.38136.17
ARC-AGI-242.4%—
NYT Connections (extended)54.5%—
ARC-AGI-187.5%—
Chess Puzzles—5%
LMArena Hard Prompts1460—
Surface Evolver Bench45%—

Math Qwen3.8 27B leads

Qwen3.8 27B: 37.1 (#161), Qwen3 8B: 34.9 (#191)

BenchmarkQwen3.8 27BQwen3 8B
OTIS Mock AIME 2024-2025—56.1%
ProofBench16%—
LMArena Math1456—

Knowledge Qwen3.8 27B leads

Qwen3.8 27B: 41.6 (#109), Qwen3 8B: 36.1 (#173)

BenchmarkQwen3.8 27BQwen3 8B
GPQA Diamond—56.8%
Vectara Hallucination Rate—4.8%
LMArena Expert1482—

Multimodal Not comparable

Qwen3.8 27B: 41.3 (#37), Qwen3 8B: —

BenchmarkQwen3.8 27BQwen3 8B
LMArena Vision1271—

Multilingual Not comparable

Qwen3.8 27B: 53.7 (#60), Qwen3 8B: —

BenchmarkQwen3.8 27BQwen3 8B
LMArena Non-English1430—
LMArena Chinese1504—
LMArena French1465—
LMArena German1438—
LMArena Japanese1384—
LMArena Korean1393—
LMArena Russian1415—
LMArena Spanish1448—

Instruction Following Not comparable

Qwen3.8 27B: 75.8 (#53), Qwen3 8B: —

BenchmarkQwen3.8 27BQwen3 8B
LMArena Instruction Following1439—

Long Context Qwen3.8 27B leads

Qwen3.8 27B: 44.3 (#70), Qwen3 8B: 37.9 (#210)

BenchmarkQwen3.8 27BQwen3 8B
Fiction.LiveBench—62.1%
LMArena Longer Query1450—

Writing & Preference Not comparable

Qwen3.8 27B: 65.8 (#43), Qwen3 8B: —

BenchmarkQwen3.8 27BQwen3 8B
LMArena Text1441—
LMArena Creative Writing1384—
EQ-Bench Creative Writing1671—
LMArena Multi-Turn1441—

Frequently asked questions

Is Qwen3.8 27B better than Qwen3 8B?

Qwen3.8 27B is the stronger model overall, scoring 46.0 to 33.7 on the Noometry Index. Qwen3 8B costs 2.0× less per token, which makes it the better buy when Qwen3.8 27B's lead doesn't matter for your workload.

Which is cheaper, Qwen3.8 27B or Qwen3 8B?

Qwen3 8B is cheaper. It lists at $0.18 per million input tokens and $0.70 per million output tokens; Qwen3.8 27B lists at $0.04 and $2.30.

Is Qwen3.8 27B or Qwen3 8B better for coding?

Qwen3.8 27B scores higher on coding benchmarks: 50.5 versus 34.0 in the Noometry coding category.

Which has the bigger context window?

Qwen3.8 27B does, with 262K tokens against 131K.

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

5 benchmarks have published results for both models. Qwen3.8 27B has 31 scored results on Noometry and Qwen3 8B has 11.

Related comparisons

Go deeper