Enter opens the highlighted result. Esc closes.

Qwen3.8 27B vs Hy4 flash

Qwen3.8 27B leads Hy4 flash on coding, 69.7 to 44.5 on the Vibe Coding Index, and Hy4 flash costs 18.1x less per output token ($0.1405 vs $2.55 per million). Updated November 22, 2026; 4 benchmarks have a result for both.

Qwen3.8 27BQwen
Coding
69.7
Overall
51.7
Output $/M
$2.55
1
Hy4 flashTencent
Coding
44.5
Overall
No data
Output $/M
$0.1405
3
vs boards won of 4 shared

Qwen3.8 27B or Hy4 flash: which to pick

Pick Qwen3.8 27B for coding: it scores 69.7 on the Vibe Coding Index against 44.5 for Hy4 flash.

Pick Hy4 flash for output tokens 18.1x cheaper ($0.1405 vs $2.55 per million) and a larger context window (1.05M against 1.02M tokens).

Shared results
4
Qwen3.8 27B only
87
Hy4 flash only
8
Like-for-like categories
2 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
Qwen3.8 27B 69.7 from 4 benchmarks, Hy4 flash 44.5 from 2 benchmarks.
Widest category
Coding, 8.6 points apart on average over 2 shared benchmarks.
Widest single gap
Vibe Code Bench (Vals AI): Qwen3.8 27B 64.8% against Hy4 flash 74.8%, 10.0 points apart on a 0 to 100 scale.
Output price
Qwen3.8 27B $2.55 against Hy4 flash $0.1405 per million output tokens, 18.1x apart.
Context window
Qwen3.8 27B takes 1.02M tokens, Hy4 flash 1.05M.

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
CategoryQwen3.8 27BBarsHy4 flashTies
Coding02
Agentic work11
All shared boards13
CodingLike-for-likeSix of these boards feed the Vibe Coding Index02
Coding benchmarks for Qwen3.8 27B and Hy4 flash
BenchmarkQwen3.8 27BHy4 flashLeader
LiveBench coding69.3%Not measuredOnly Qwen3.8 27B measured
LMArena WebDev1590Not measuredOnly Qwen3.8 27B measured
LiveCodeBench (Vals AI)84.0%Not measuredOnly Qwen3.8 27B measured
SWE-bench Verified (Vals AI)86.0%Not measuredOnly Qwen3.8 27B measured
Terminal-Bench 4 (Vals AI)Not measured15.2%Only Hy4 flash measured
Vibe Code Bench (Vals AI)64.8%74.8%Hy4 flash
Code migration (Vals AI)14.2%Not measuredOnly Qwen3.8 27B measured
ProgramBench (Vals AI)0.0%Not measuredOnly Qwen3.8 27B measured
IOI (Vals AI)39.1%Not measuredOnly Qwen3.8 27B measured
Terminal-Bench 2.1 (Vals AI)58.4%Not measuredOnly Qwen3.8 27B measured
SkillsBench (Vals AI)38.1%Not measuredOnly Qwen3.8 27B measured
CyberBench (Vals AI)67.8%Not measuredOnly Qwen3.8 27B measured
SciCode46.6%Not measuredOnly Qwen3.8 27B measured
Terminal-Bench 4.0 (Artificial Analysis run)5.6%Not measuredOnly Qwen3.8 27B measured
FrontierCode 2.014.9%Not measuredOnly Qwen3.8 27B measured
LMArena Text: coding1481Not measuredOnly Qwen3.8 27B measured
LMArena Text: software and IT1475Not measuredOnly Qwen3.8 27B measured
LMArena WebDev: React1594Not measuredOnly Qwen3.8 27B measured
LMArena WebDev: HTML1571Not measuredOnly Qwen3.8 27B measured
LMArena WebDev: from an image1575Not measuredOnly Qwen3.8 27B measured
GameBench 243.450.5Hy4 flash
Neither model has a result on the other 31 boards here
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
Intelligence benchmarks for Qwen3.8 27B and Hy4 flash
BenchmarkQwen3.8 27BHy4 flashLeader
AA Intelligence Index20.1Not measuredOnly Qwen3.8 27B measured
Epoch Capabilities Index152.6Not measuredOnly Qwen3.8 27B measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index11
Agentic work benchmarks for Qwen3.8 27B and Hy4 flash
BenchmarkQwen3.8 27BHy4 flashLeader
LiveBench research61.3%Not measuredOnly Qwen3.8 27B measured
LiveBench browser24.2%Not measuredOnly Qwen3.8 27B measured
LiveBench long agentic15.9%Not measuredOnly Qwen3.8 27B measured
Finance Agent (Vals AI)48.5%Not measuredOnly Qwen3.8 27B measured
Tax Agent Bench (Vals AI)30.9%23.3%Qwen3.8 27B
Legal Research Bench (Vals AI)36.1%38.5%Hy4 flash
Harvey's Legal Agent Benchmark (Vals AI)Not measured5.4%Only Hy4 flash measured
Excel Modeling Benchmark (Vals AI)59.7%Not measuredOnly Qwen3.8 27B measured
Terminal-Bench-Science (Vals AI)0.0%Not measuredOnly Qwen3.8 27B measured
APEX-Agents47.5%Not measuredOnly Qwen3.8 27B measured
Neither model has a result on the other 34 boards here
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index15 boards, none shared
Reasoning benchmarks for Qwen3.8 27B and Hy4 flash
BenchmarkQwen3.8 27BHy4 flashLeader
GPQA Diamond88.9%Not measuredOnly Qwen3.8 27B measured
LiveBench reasoning80.0%Not measuredOnly Qwen3.8 27B measured
MMMU Pro (Vals AI)83.9%Not measuredOnly Qwen3.8 27B measured
GPQA Diamond (Artificial Analysis run)90.5%Not measuredOnly Qwen3.8 27B measured
Humanity's Last Exam (Artificial Analysis run)33.9%Not measuredOnly Qwen3.8 27B measured
HLE-Diamond (Artificial Analysis run)15.3%Not measuredOnly Qwen3.8 27B measured
CritPt v2, hard subset (Artificial Analysis run)16.5%Not measuredOnly Qwen3.8 27B measured
ARC-AGI87.5%Not measuredOnly Qwen3.8 27B measured
ARC-AGI-242.4%Not measuredOnly Qwen3.8 27B measured
SimpleBench60.2%Not measuredOnly Qwen3.8 27B measured
LMArena Text: hard prompts1458Not measuredOnly Qwen3.8 27B measured
LMArena Text: expert1479Not measuredOnly Qwen3.8 27B measured
LMArena Text: hard prompts in English1463Not measuredOnly Qwen3.8 27B measured
LMArena Vision: diagrams1280Not measuredOnly Qwen3.8 27B measured
LMArena Vision: homework1288Not measuredOnly Qwen3.8 27B measured
Neither model has a result on the other 11 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index3 boards, none shared
Math benchmarks for Qwen3.8 27B and Hy4 flash
BenchmarkQwen3.8 27BHy4 flashLeader
ProofBench16.0%Not measuredOnly Qwen3.8 27B measured
LMArena Text: math1459Not measuredOnly Qwen3.8 27B measured
LMArena Text: mathematical work1462Not measuredOnly Qwen3.8 27B measured
Neither model has a result on the other 11 boards here
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index11 boards, none shared
Knowledge benchmarks for Qwen3.8 27B and Hy4 flash
BenchmarkQwen3.8 27BHy4 flashLeader
MMLU Pro84.3%Not measuredOnly Qwen3.8 27B measured
LegalBench (Vals AI)82.4%Not measuredOnly Qwen3.8 27B measured
TaxEval (Vals AI)70.8%Not measuredOnly Qwen3.8 27B measured
MortgageTax (Vals AI)64.9%Not measuredOnly Qwen3.8 27B measured
MedCode (Vals AI)28.7%Not measuredOnly Qwen3.8 27B measured
AA-Omniscience-6.4Not measuredOnly Qwen3.8 27B measured
GeoBench38.0%Not measuredOnly Qwen3.8 27B measured
LMArena Text: science1465Not measuredOnly Qwen3.8 27B measured
LMArena Text: medicine and healthcare1439Not measuredOnly Qwen3.8 27B measured
LMArena Text: legal and government1447Not measuredOnly Qwen3.8 27B measured
LMArena Text: business and finance1440Not measuredOnly Qwen3.8 27B measured
Neither model has a result on the other 10 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
Web design benchmarks for Qwen3.8 27B and Hy4 flash
BenchmarkQwen3.8 27BHy4 flashLeader
Design Arena: websitesNot measured1227Only Hy4 flash measured
Design Arena: UI componentsNot measured1262Only Hy4 flash measured
Design Arena: game devNot measured1202Only Hy4 flash measured
Design Arena: 3DNot measured1209Only Hy4 flash measured
Design Arena: all code categoriesNot measured1225Only Hy4 flash measured
LiveBench design72.4%Not measuredOnly Qwen3.8 27B measured
Neither model has a result on the other 3 boards here
GeneralDirectional onlyPublic boards, shown as published, kept out of the index31 boards, none shared
General benchmarks for Qwen3.8 27B and Hy4 flash
BenchmarkQwen3.8 27BHy4 flashLeader
Vals IndexNot measured50.8%Only Hy4 flash measured
LiveBench data analysis76.6%Not measuredOnly Qwen3.8 27B measured
MedScribe (Vals AI)83.8%Not measuredOnly Qwen3.8 27B measured
Long context reasoning (AA-LCR)82.0%Not measuredOnly Qwen3.8 27B measured
GDPval-AA1426Not measuredOnly Qwen3.8 27B measured
DTBench88.0%Not measuredOnly Qwen3.8 27B measured
LMCA41.4%Not measuredOnly Qwen3.8 27B measured
Surface Evolver Bench45.0%Not measuredOnly Qwen3.8 27B measured
LMArena Text1437Not measuredOnly Qwen3.8 27B measured
LMArena Text: instruction following1433Not measuredOnly Qwen3.8 27B measured
LMArena Text: creative writing1378Not measuredOnly Qwen3.8 27B measured
LMArena Vision1271Not measuredOnly Qwen3.8 27B measured
LMArena Text: English1448Not measuredOnly Qwen3.8 27B measured
LMArena Text: non-English1425Not measuredOnly Qwen3.8 27B measured
LMArena Text: Chinese1491Not measuredOnly Qwen3.8 27B measured
LMArena Text: German1437Not measuredOnly Qwen3.8 27B measured
LMArena Text: Japanese1390Not measuredOnly Qwen3.8 27B measured
LMArena Text: Korean1398Not measuredOnly Qwen3.8 27B measured
LMArena Text: French1466Not measuredOnly Qwen3.8 27B measured
LMArena Text: Spanish1445Not measuredOnly Qwen3.8 27B measured
LMArena Text: Russian1412Not measuredOnly Qwen3.8 27B measured
LMArena Text: Polish1449Not measuredOnly Qwen3.8 27B measured
LMArena Text: multi-turn1439Not measuredOnly Qwen3.8 27B measured
LMArena Text: longer queries1446Not measuredOnly Qwen3.8 27B measured
LMArena Text: writing and language1402Not measuredOnly Qwen3.8 27B measured
LMArena Text: entertainment and media1389Not measuredOnly Qwen3.8 27B measured
LMArena Vision: English1275Not measuredOnly Qwen3.8 27B measured
LMArena Vision: Chinese1327Not measuredOnly Qwen3.8 27B measured
LMArena Vision: reading text1279Not measuredOnly Qwen3.8 27B measured
LMArena Vision: humor1242Not measuredOnly Qwen3.8 27B measured
LMArena Vision: creative writing1238Not measuredOnly Qwen3.8 27B measured
Neither model has a result on the other 18 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Qwen3.8 27B and Hy4 flash
BenchmarkQwen3.8 27BHy4 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

Qwen3.8 27B$0.0017Fits in one request
Hy4 flash$0.0001Fits in one request

Hy4 flash costs less

Repository review

50K input + 3K output tokens

Qwen3.8 27B$0.029Fits in one request
Hy4 flash$0.0022Fits in one request

Hy4 flash costs less

Agent loop

200K input + 10K output tokens

Qwen3.8 27B$0.111Fits in one request
Hy4 flash$0.0084Fits in one request

Hy4 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 Qwen3.8 27B and Hy4 flash
SpecificationQwen3.8 27BTencent Hy4 flash
MakerQwenTencent
WeightsOpen weightsOpen weights
ReleasedAugust 14, 2026November 16, 2026
Context window1.02M1.05M
Input $/Mtok$0.425$0.035
Output $/Mtok$2.55$0.1405
Coding index69.744.5
Independent measurements behind it42
Cost per solved task$1.26Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Qwen3.8 27B better than Hy4 flash for coding?

Pick Qwen3.8 27B for coding: it scores 69.7 on the Vibe Coding Index against 44.5 for Hy4 flash. Pick Hy4 flash for output tokens 18.1x cheaper ($0.1405 vs $2.55 per million) and a larger context window (1.05M against 1.02M tokens).

Which is cheaper, Qwen3.8 27B or Hy4 flash?

Qwen3.8 27B costs $0.425 per million input tokens and $2.55 per million output tokens; Hy4 flash costs $0.035 and $0.1405. Cost per solved task is known for Qwen3.8 27B only.

Which has the bigger context window, Qwen3.8 27B or Hy4 flash?

Hy4 flash, with 1.05M tokens against 1.02M.

How many benchmarks is the Qwen3.8 27B vs Hy4 flash comparison based on?

4 benchmarks have a result for both models as of November 22, 2026. Qwen3.8 27B's coding index rests on 4 independent benchmarks and Hy4 flash's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/qwen3-8-27b-vs-tencent-hy4-flash" title="Qwen3.8 27B vs Hy4 flash, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/qwen3-8-27b-vs-tencent-hy4-flash">LLMBenchmark.ai, Qwen3.8 27B vs Hy4 flash</a></p>