LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

o3 vs Qwen3.8 27B

Updated October 21, 2026. 55 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace o3 Replace Qwen3.8 27B
o3OpenAI
Coding
70.7
Overall
20.2
Output $/M
$8.00
7
Qwen3.8 27BQwen
Coding
73.6
Overall
33.7
Output $/M
$1.35
48
vs boards won of 55 shared

Decision reading

Pick Qwen3.8 27B for coding: it scores 73.6 on the Vibe Coding Index against 70.7 for o3.

Pick o3 when a solved task costs $0.571 against $1.26, or when what matters is MMLU Pro or LegalBench (Vals AI), where it leads.

Model Ao3OpenAI70.7/1002 independent benchmarks$0.571 per solved taskModel BQwen3.8 27BQwen73.6/1002 independent benchmarks$1.26 per solved task
Shared results
55
o3 only
42
Qwen3.8 27B only
30
Like-for-like categories
6 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

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
o3 70.7 from 2 benchmarks, Qwen3.8 27B 73.6 from 2 benchmarks.
Widest category
Reasoning, 13.1 points apart on average over 11 shared benchmarks.
Widest single gap
ARC-AGI-2: o3 6.5% against Qwen3.8 27B 42.4%, 35.8 points apart on a 0 to 100 scale.
Output price
o3 $8.00 against Qwen3.8 27B $1.35 per million output tokens, 5.9x apart.
Context window
o3 takes 200K tokens, Qwen3.8 27B 1M.

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
Categoryo3BarsQwen3.8 27BTies
Coding03
Intelligence02
Reasoning011
Math02
Knowledge63
General127
All shared boards748
CodingLike-for-likeFour of these boards feed the Vibe Coding Index03
Coding benchmarks for o3 and Qwen3.8 27B
Benchmarko3Qwen3.8 27BLeader
Aider polyglot81.3%Not measuredOnly o3 measured
SWE-bench Verified58.4%Not measuredOnly o3 measured
SWE-bench Multimodal36.0%Not measuredOnly o3 measured
LiveBench codingNot measured86.8%Only Qwen3.8 27B measured
LMArena WebDevNot measured1593Only Qwen3.8 27B measured
AA Coding IndexNot measured68.1Only Qwen3.8 27B measured
LiveCodeBench (Vals AI)83.9%84.0%Qwen3.8 27B
SWE-bench Verified (Vals AI)Not measured86.0%Only Qwen3.8 27B measured
Vibe Code Bench (Vals AI)Not measured64.8%Only Qwen3.8 27B measured
Code migration (Vals AI)Not measured14.2%Only Qwen3.8 27B measured
ProgramBench (Vals AI)Not measured0.0%Only Qwen3.8 27B measured
IOI (Vals AI)Not measured39.1%Only Qwen3.8 27B measured
Terminal-Bench 2.1 (Vals AI)Not measured58.4%Only Qwen3.8 27B measured
SkillsBench (Vals AI)Not measured38.1%Only Qwen3.8 27B measured
CyberBench (Vals AI)Not measured67.8%Only Qwen3.8 27B measured
LiveCodeBench (Artificial Analysis run)80.8%Not measuredOnly o3 measured
SciCodeNot measured46.6%Only Qwen3.8 27B measured
Terminal-Bench Hard (Artificial Analysis run)37.1%Not measuredOnly o3 measured
GSO-Bench8.8%Not measuredOnly o3 measured
WeirdML52.4%Not measuredOnly o3 measured
ALE-Bench933.5Not measuredOnly o3 measured
LMArena Text: coding14071483Qwen3.8 27B
LMArena Text: software and IT14151478Qwen3.8 27B
LMArena WebDev: ReactNot measured1602Only Qwen3.8 27B measured
LMArena WebDev: HTMLNot measured1570Only Qwen3.8 27B measured
LMArena WebDev: from an imageNot measured1577Only Qwen3.8 27B measured
Neither model has a result on the other 12 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for o3 and Qwen3.8 27B
Benchmarko3Qwen3.8 27BLeader
AA Intelligence Index20.233.7Qwen3.8 27B
Epoch Capabilities Index146.7151.7Qwen3.8 27B
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index15 boards, none shared
Agentic work benchmarks for o3 and Qwen3.8 27B
Benchmarko3Qwen3.8 27BLeader
AA Agentic IndexNot measured45.8Only Qwen3.8 27B measured
LiveBench researchNot measured59.4%Only Qwen3.8 27B measured
LiveBench browserNot measured32.0%Only Qwen3.8 27B measured
LiveBench long agenticNot measured27.9%Only Qwen3.8 27B measured
Finance Agent (Vals AI)Not measured48.5%Only Qwen3.8 27B measured
Tax Agent Bench (Vals AI)Not measured30.9%Only Qwen3.8 27B measured
Legal Research Bench (Vals AI)Not measured36.1%Only Qwen3.8 27B measured
Excel Modeling Benchmark (Vals AI)Not measured59.7%Only Qwen3.8 27B measured
Terminal-Bench-Science (Vals AI)Not measured0.0%Only Qwen3.8 27B measured
τ²-Bench Telecom80.7%Not measuredOnly o3 measured
APEX-AgentsNot measured47.5%Only Qwen3.8 27B measured
METR time horizons65.4%Not measuredOnly o3 measured
GDPval30.8%Not measuredOnly o3 measured
DeepResearch Bench45.2%Not measuredOnly o3 measured
OSWorld23.0%Not measuredOnly o3 measured
Neither model has a result on the other 24 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index011
Reasoning benchmarks for o3 and Qwen3.8 27B
Benchmarko3Qwen3.8 27BLeader
GPQA Diamond84.1%88.9%Qwen3.8 27B
LiveBench reasoningNot measured80.0%Only Qwen3.8 27B measured
MMMU Pro (Vals AI)80.4%83.9%Qwen3.8 27B
GPQA Diamond (Artificial Analysis run)82.7%90.5%Qwen3.8 27B
Humanity's Last Exam (Artificial Analysis run)20.1%33.9%Qwen3.8 27B
GPQA Diamond (Epoch AI run)81.8%Not measuredOnly o3 measured
ARC-AGI60.8%87.5%Qwen3.8 27B
ARC-AGI-26.5%42.4%Qwen3.8 27B
SimpleBench53.1%Not measuredOnly o3 measured
Chess puzzles37.0%Not measuredOnly o3 measured
Mystery game puzzles29.0%Not measuredOnly o3 measured
VPCT52.0%Not measuredOnly o3 measured
LMArena Text: hard prompts14021459Qwen3.8 27B
LMArena Text: expert14021479Qwen3.8 27B
LMArena Text: hard prompts in English14111461Qwen3.8 27B
LMArena Vision: diagrams12201285Qwen3.8 27B
LMArena Vision: homework12341295Qwen3.8 27B
Neither model has a result on the other 4 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index02
Math benchmarks for o3 and Qwen3.8 27B
Benchmarko3Qwen3.8 27BLeader
AIME85.3%Not measuredOnly o3 measured
MATH 500 (Vals AI)94.6%Not measuredOnly o3 measured
MGSM (Vals AI)91.8%Not measuredOnly o3 measured
AA Math Index88.3Not measuredOnly o3 measured
AIME 2025 (Artificial Analysis run)88.3%Not measuredOnly o3 measured
OTIS Mock AIME84.4%Not measuredOnly o3 measured
FrontierMath, tiers 1 to 333.3%Not measuredOnly o3 measured
ProofBenchNot measured16.0%Only Qwen3.8 27B measured
MATH level 597.8%Not measuredOnly o3 measured
LMArena Text: math14261455Qwen3.8 27B
LMArena Text: mathematical work14231458Qwen3.8 27B
Neither model has a result on the other 2 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index63
Knowledge benchmarks for o3 and Qwen3.8 27B
Benchmarko3Qwen3.8 27BLeader
MMLU Pro85.6%84.3%o3
LegalBench (Vals AI)83.8%82.4%o3
CorpFin (Vals AI)59.7%Not measuredOnly o3 measured
TaxEval (Vals AI)74.6%70.8%o3
MortgageTax (Vals AI)65.7%64.9%o3
MedQA (Vals AI)96.1%Not measuredOnly o3 measured
MedCode (Vals AI)47.3%28.7%o3
MMLU Pro (Artificial Analysis run)85.3%Not measuredOnly o3 measured
SimpleQA Verified49.4%Not measuredOnly o3 measured
GeoBench74.0%Not measuredOnly o3 measured
LMArena Text: science14291467Qwen3.8 27B
LMArena Text: medicine and healthcare14481437o3
LMArena Text: legal and government14281447Qwen3.8 27B
LMArena Text: business and finance13931438Qwen3.8 27B
Neither model has a result on the other 3 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
Web design benchmarks for o3 and Qwen3.8 27B
Benchmarko3Qwen3.8 27BLeader
Design Arena: websites1036Not measuredOnly o3 measured
Design Arena: UI components1016Not measuredOnly o3 measured
Design Arena: game dev1034Not measuredOnly o3 measured
Design Arena: data viz1200Not measuredOnly o3 measured
Design Arena: all code categories1021Not measuredOnly o3 measured
LiveBench designNot measured75.9%Only Qwen3.8 27B measured
Neither model has a result on the other 3 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index127
General benchmarks for o3 and Qwen3.8 27B
Benchmarko3Qwen3.8 27BLeader
LiveBench data analysisNot measured76.6%Only Qwen3.8 27B measured
SAGE (Vals AI)41.8%52.4%Qwen3.8 27B
MedScribe (Vals AI)76.7%83.8%Qwen3.8 27B
IFBench71.4%Not measuredOnly o3 measured
Long context reasoning (AA-LCR)74.7%82.0%Qwen3.8 27B
DTBench84.8%88.0%Qwen3.8 27B
LMCA39.7%41.4%Qwen3.8 27B
Fiction.LiveBench88.9%Not measuredOnly o3 measured
Creative writing (Lech Mazur)83.9%Not measuredOnly o3 measured
CL-bench17.8%Not measuredOnly o3 measured
Surface Evolver BenchNot measured45.0%Only Qwen3.8 27B measured
LMArena Text14101440Qwen3.8 27B
LMArena Text: instruction following13681436Qwen3.8 27B
LMArena Text: creative writing13591383Qwen3.8 27B
LMArena Vision12141269Qwen3.8 27B
LMArena Text: English14171448Qwen3.8 27B
LMArena Text: non-English14011428Qwen3.8 27B
LMArena Text: Chinese14371501Qwen3.8 27B
LMArena Text: German14201443Qwen3.8 27B
LMArena Text: Japanese14031387o3
LMArena Text: Korean13701384Qwen3.8 27B
LMArena Text: French14301465Qwen3.8 27B
LMArena Text: Spanish13951449Qwen3.8 27B
LMArena Text: Russian14061415Qwen3.8 27B
LMArena Text: Polish14381439Qwen3.8 27B
LMArena Text: multi-turn14051440Qwen3.8 27B
LMArena Text: longer queries13721448Qwen3.8 27B
LMArena Text: writing and language13721405Qwen3.8 27B
LMArena Text: entertainment and media13731393Qwen3.8 27B
LMArena Vision: English12211267Qwen3.8 27B
LMArena Vision: Chinese12371326Qwen3.8 27B
LMArena Vision: reading text12181281Qwen3.8 27B
LMArena Vision: humor12221235Qwen3.8 27B
LMArena Vision: captioning1210Not measuredOnly o3 measured
LMArena Vision: recognizing things1234Not measuredOnly o3 measured
LMArena Vision: creative writing12011237Qwen3.8 27B
Neither model has a result on the other 10 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for o3 and Qwen3.8 27B
Benchmarko3Qwen3.8 27BLeader
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

o3$0.0060Fits in one request
Qwen3.8 27B$0.0007Fits in one request

Qwen3.8 27B costs less

Repository review

50K input + 3K output tokens

o3$0.124Fits in one request
Qwen3.8 27B$0.0060Fits in one request

Qwen3.8 27B costs less

Agent loop

200K input + 10K output tokens

o3$0.480Exceeds the context window
Qwen3.8 27B$0.021Fits in one request

Qwen3.8 27B 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 o3 and Qwen3.8 27B
SpecificationOpenAI o3Qwen3.8 27B
MakerOpenAIQwen
WeightsNo public weights listedOpen weights
ReleasedApril 16, 2025August 14, 2026
Context window200K1M
Input $/Mtok$2.00$0.04
Output $/Mtok$8.00$1.35
Coding index70.773.6
Independent measurements behind it22
Cost per solved task$0.571$1.26
SourceOpenRouter listingOpenRouter listing

Questions

Is o3 better than Qwen3.8 27B for coding?

Pick Qwen3.8 27B for coding: it scores 73.6 on the Vibe Coding Index against 70.7 for o3. Pick o3 when a solved task costs $0.571 against $1.26, or when what matters is MMLU Pro or LegalBench (Vals AI), where it leads.

Which is cheaper, o3 or Qwen3.8 27B?

o3 costs $2.00 per million input tokens and $8.00 per million output tokens; Qwen3.8 27B costs $0.04 and $1.35. Per solved SWE-bench Verified task, o3 costs $0.571 and Qwen3.8 27B $1.26.

Which has the bigger context window, o3 or Qwen3.8 27B?

Qwen3.8 27B, with 1M tokens against 200K.

How many benchmarks is the o3 vs Qwen3.8 27B comparison based on?

55 benchmarks have a result for both models as of October 21, 2026. o3's coding index rests on 2 independent benchmarks and Qwen3.8 27B'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/o3-vs-qwen3-8-27b" title="o3 vs Qwen3.8 27B, 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/o3-vs-qwen3-8-27b">LLMBenchmark.ai, o3 vs Qwen3.8 27B</a></p>