GPT-6 Lunavs.Qwen 3.8 27B

Which one is better?

GPT-6 Luna is clearly ahead in the overall index. Qwen 3.8 27B is cheaper for input, GPT-6 Luna for output.

IntelligenceOverall index, 0 to 100 · higher is better

GPT-6 Luna

Rank 31 to 39

Qwen 3.8 27B

Rank 44 to 57

SpeedTokens per second · higher is better

122

GPT-6 Luna

fairly fast

Qwen 3.8 27B

slow

PriceDollars per 1M tokens · lower is better

$0.20

GPT-6 Luna

cheap

$0.66

Qwen 3.8 27B

fairly cheap

GPT-6 Luna

Subscription: Free to use at OpenAI (ChatGPT Free), more from €8 a month (ChatGPT Go). Through other providers from ≈ €1.06 a month (Command Code Go, with limited credits). 57 plans in Germany.
Via API: $0.10 input, $0.50 output per 1M tokens, unchanged since Sep 24, 2026.

Qwen 3.8 27B

Subscription: No Alibaba subscription includes this model. Through other providers from ≈ €1.06 a month (Command Code Go, with limited credits). 10 plans in Germany.
Via API: $0.04 input, $2.50 output per 1M tokens according to the provider; the price index has listed it at $0.42 and $3 since Oct 27, 2026.

Overall index with data through Oct 29, 2026 · Prices from Oct 29, 2026 · Speed: measured by Artificial Analysis · Subscriptions as of Oct 24, 2026 · Subscription prices incl. VAT

Where is the difference?

Values from 0 to 100 per task area. We only name a winner when the gap is large enough and there are enough tests.

GPT-6 Luna ahead: 7Qwen 3.8 27B ahead: 0level: 2too little data: 1of 10 areas

AreaGPT-6 LunaQwen 3.8 27B0 to 100, higher is betterResult

Areas in the overall index

CodingGPT-6 Luna ahead4413

ReasoningGPT-6 Luna ahead466

MathematicsGPT-6 Luna ahead668

KnowledgeGPT-6 Luna ahead185

AgentsGPT-6 Luna ahead4821

Generalonly GPT-6 Luna has a score35not enough data

MultimodalGPT-6 Luna ahead2413

Specialist fields

Financelevel4534

Legallevel2423

HealthGPT-6 Luna ahead353

Show 4 more areas that do not count toward the overall index

AreaGPT-6 LunaQwen 3.8 27B0 to 100, higher is betterResult

CybersecurityGPT-6 Luna ahead6619

Searchonly GPT-6 Luna has a score4not enough data

Productivityonly GPT-6 Luna has a score19not enough data

Educationonly GPT-6 Luna has a score27not enough data

See for yourself: same task, one attempt

The same prompt for both models. We show the latest attempt without fixes; it does not count toward the overall index.

Pelican on a bicycle

A community classic for free-form SVG drawing.

GPT-6 Luna

Reasoning (requested): high

Current attempt

Run date: Oct 1, 2026
API provider: openrouter
Requested API model: openai/gpt-6-luna
API model reported in response: openai/gpt-6-luna
Reasoning setting: high
Pinned endpoint: OpenAI | openai/gpt-6-luna-20260922
Temperature: Provider default
Token limit including reasoning: 8,192 tokens
Test protocol: community-visual-tests-v1
Technical checks: Passed, not a quality rating
Response time: 46.7 s

Visually reviewed. This individual test does not establish overall model quality.

Qwen 3.8 27B

Not tested yet

Prompt and test conditions

Generate an SVG of a pelican riding a bicycle

Token limit including reasoning: 8,192 tokens

Task origin (Simon Willison)

SVG short test

One SVG task with the same fixed settings for every model, without hidden thinking.

Shared test round: 43b8c5c46830

GPT-6 Luna

Unavailable in this round

Qwen 3.8 27B

All data side by side

The better value is bold. For the overall index only when the gap is clear.

Metric	GPT-6 Luna	Qwen 3.8 27B
Intelligence
Overall index0 to 100, higher is better	38	11
Rank	Rank 31 to 39 of 95	Rank 44 to 57 of 95
Price per 1M tokens
Inputwhat you send	$0.10above 272,000 tokens: $0.20	$0.04
Outputwhat the model writes	$0.50above 272,000 tokens: $0.75	$2.50
Speed
Tokens per second	122	40
Technical
Context windowhow much text the model reads at once	1.05 M tokens≈ 1,580 A4 pages	262,144 tokens≈ 390 A4 pages
Knowledge cutoff	May 18, 2026	n/a
Released	Sep 22, 2026	Aug 14, 2026
Availability	Active	Active
Open weights	no	yes

Prices without caching or batch discounts · Page count as a rule of thumb from Artificial Analysis

Shared benchmarks

Single results from the same benchmark version, task and metric. A value is bold only with the same test setup and a clear gap.

A: GPT-6 LunaB: Qwen 3.8 27B

AA-BriefcaseOverall · higher is better1,234.65 elo1,392.67 eloArtificial AnalysisAs of 10/29/2026

AA-OmniscienceOverall · higher is better-12.967 points-6.367 pointsArtificial AnalysisAs of 10/29/2026

AA-Omniscience AccuracyOverall · higher is better38.6%17.3%Artificial AnalysisAs of 10/29/2026

AA-Omniscience Hallucination RateOverall · lower is better84.0%28.6%Artificial AnalysisAs of 10/29/2026

Artificial Analysis Intelligence IndexOverall · higher is better23.301 points20.107 pointsArtificial AnalysisAs of 10/29/2026

AutomationBench-AAOverall · higher is better47.8%48.2%Artificial AnalysisAs of 10/29/2026

Code MigrationOverall · higher is betterTest setup not documented as identical42.6%14.2%Vals AIAs of 10/18/2026

CritPt v2Overall · higher is better17.4%5.4%Artificial AnalysisAs of 10/29/2026

CritPt v2 (Hard Subset)Overall · higher is better35.5%4.0%Artificial AnalysisAs of 10/29/2026

CyberBench v1.1Overall · higher is betterTest setup not documented as identical76.3%67.8%Vals AIAs of 10/20/2026

Show 54 more benchmarks

EMBOverall · higher is betterTest setup not documented as identical68.5%59.7%Vals AIAs of 10/20/2026

Finance Agent (v2)Overall · higher is betterTest setup not documented as identical49.9%48.6%Vals AIAs of 10/20/2026

FrontierCodeOverall · higher is better18.4%14.9%Artificial AnalysisAs of 10/29/2026

FrontierCode 1.1 MainOverall · higher is better33.2%34.2%Artificial AnalysisAs of 10/29/2026

GDP.pdfOverall · higher is better18.6%16.4%Artificial AnalysisAs of 10/29/2026

GDPval-AAOverall · higher is better1,373.25 elo1,427.26 eloArtificial AnalysisAs of 10/29/2026

HLE-DiamondOverall · higher is better18.9%15.3%Artificial AnalysisAs of 10/29/2026

IOIOverall · higher is betterTest setup not documented as identical55.6%39.1%Vals AIAs of 10/24/2026

Legal Research BenchOverall · All-pass · higher is betterTest setup not documented as identical30.3%36.1%Vals AIAs of 10/26/2026

MedCodeOverall · higher is betterTest setup not documented as identical44.7%28.7%Vals AIAs of 10/13/2026

MedScribeOverall · higher is betterTest setup not documented as identical83.7%83.8%Vals AIAs of 10/13/2026

MMMU-ProOverall · higher is better79.7%76.3%Artificial AnalysisAs of 10/29/2026

ProgramBenchOverall · higher is betterTest setup not documented as identical0.5%0%Vals AIAs of 10/20/2026

ProofBench v1.1Overall · higher is betterTest setup not documented as identical64%16%Vals AIAs of 10/24/2026

Tax Agent BenchOverall · higher is betterTest setup not documented as identical19.9%30.9%Vals AIAs of 10/26/2026

Terminal-BenchOverall · higher is better8.1%5.6%Artificial AnalysisAs of 10/29/2026

Terminal-Bench 2.1Overall · higher is betterTest setup not documented as identical73%58.4%Vals AIAs of 10/01/2026

Terminal-Bench-ScienceOverall · higher is betterTest setup not documented as identical4.3%0%Vals AIAs of 10/25/2026

Terminal-Bench-ScienceOverall · higher is better5.2%0.5%Artificial AnalysisAs of 10/29/2026

Vibe Code Bench v1.1Overall · higher is betterTest setup not documented as identical81.6%64.9%Vals AIAs of 10/26/2026

AA-Omniscience AccuracyOverall · higher is better38.7%17.9%Artificial AnalysisAs of 10/29/2026

AA-Omniscience Hallucination RateOverall · lower is better85.3%65.4%Artificial AnalysisAs of 10/29/2026

Code MigrationCLI · higher is betterTest setup not documented as identical37.2%2.2%Vals AIAs of 10/18/2026

CritPt v2Overall · higher is better10.6%0.0%Artificial AnalysisAs of 10/29/2026

CyberBench v1.1Patch · higher is betterTest setup not documented as identical87.5%83.9%Vals AIAs of 10/20/2026

EMBComps · higher is betterTest setup not documented as identical70.5%53.3%Vals AIAs of 10/20/2026

Finance Agent (v2)Adjustments · higher is betterTest setup not documented as identical49.8%47.3%Vals AIAs of 10/20/2026

FrontierCode 1.1 MainOverall · higher is better31.8%34.7%Artificial AnalysisAs of 10/29/2026

GDPval-AAOverall · higher is better1,270.95 elo1,389.29 eloArtificial AnalysisAs of 10/29/2026

IOIIOI 2024 · higher is betterTest setup not documented as identical69.3%31.8%Vals AIAs of 10/24/2026

Legal Research BenchAdministrative / Regulatory · All-pass · higher is betterTest setup not documented as identical37.9%45.5%Vals AIAs of 10/26/2026

MMMU-ProOverall · higher is better77.9%74.2%Artificial AnalysisAs of 10/29/2026

ProgramBenchAlmost Resolved · higher is betterTest setup not documented as identical8%1%Vals AIAs of 10/20/2026

Tax Agent BenchControversy & Precedence Analysis · higher is betterTest setup not documented as identical6.7%6.7%Vals AIAs of 10/26/2026

Terminal-Bench 2.1Easy · higher is betterTest setup not documented as identical91.7%50%Vals AIAs of 10/01/2026

Terminal-Bench-ScienceEarth Sciences · higher is betterTest setup not documented as identical0%0%Vals AIAs of 10/25/2026

AA-Omniscience AccuracyOverall · higher is better30.1%9.1%Artificial AnalysisAs of 10/29/2026

AA-Omniscience Hallucination RateOverall · lower is better80.0%18.0%Artificial AnalysisAs of 10/29/2026

Code MigrationCOBOL · higher is betterTest setup not documented as identical58.6%50%Vals AIAs of 10/18/2026

CritPt v2Overall · higher is better1.1%0.3%Artificial AnalysisAs of 10/29/2026

CyberBench v1.1PoC · higher is betterTest setup not documented as identical65%51.7%Vals AIAs of 10/20/2026

EMBDataroom · higher is betterTest setup not documented as identical75.5%79.7%Vals AIAs of 10/20/2026

Finance Agent (v2)All-Pass · higher is betterTest setup not documented as identical38.4%36.8%Vals AIAs of 10/20/2026

FrontierCode 1.1 MainOverall · higher is better13.8%7.9%Artificial AnalysisAs of 10/29/2026

GDPval-AAOverall · higher is better1,052.09 elo1,082.47 eloArtificial AnalysisAs of 10/29/2026

IOIIOI 2025 · higher is betterTest setup not documented as identical39%48.5%Vals AIAs of 10/24/2026

Legal Research BenchAdministrative / Regulatory · Weighted partial credit · higher is betterTest setup not documented as identical83.2%88.5%Vals AIAs of 10/26/2026

MMMU-ProOverall · higher is better62.1%69.9%Artificial AnalysisAs of 10/29/2026

ProgramBenchFully Resolved · higher is betterTest setup not documented as identical0.5%0%Vals AIAs of 10/20/2026

Tax Agent BenchControversy & Precedence Analysis · higher is betterTest setup not documented as identical54.4%48.3%Vals AIAs of 10/26/2026

Terminal-Bench 2.1Hard · higher is betterTest setup not documented as identical65.6%43.3%Vals AIAs of 10/01/2026

Terminal-Bench-ScienceEngineering Sciences · higher is betterTest setup not documented as identical11.1%0%Vals AIAs of 10/25/2026

AA-Omniscience AccuracyOverall · higher is better36.1%18.3%Artificial AnalysisAs of 10/29/2026

AA-Omniscience Hallucination RateOverall · lower is better84.8%48.7%Artificial AnalysisAs of 10/29/2026

Compare another matchup

First modelSecond model

More matchups

Compare GPT-6 Luna and Qwen 3.8 27B with other leading models.

View the complete LLM comparison

GPT-6 Lunavs.Qwen 3.8 27B

GPT-6 Luna

Qwen 3.8 27B

Where is the difference?

See for yourself: same task, one attempt

Pelican on a bicycle

SVG short test

GPT-6 Luna

Qwen 3.8 27B

All data side by side

Shared benchmarks

Compare another matchup

Related LLM comparisons