Enter opens the highlighted result. Esc closes.

Gemini 3.8 Flash vs GPT-6 Luna

Gemini 3.8 Flash leads GPT-6 Luna on coding, 61.6 to 57.3 on the Vibe Coding Index, and GPT-6 Luna costs 7.5x less per output token ($0.50 vs $3.75 per million). Updated October 22, 2026; 102 benchmarks have a result for both.

Replace Gemini 3.8 Flash Replace GPT-6 Luna
Gemini 3.8 FlashGoogle
Coding
61.6
Overall
69.0
Output $/M
$3.75
86
GPT-6 LunaOpenAI
Coding
57.3
Overall
59.7
Output $/M
$0.50
15
vs boards won of 102 shared, 1 tied

Gemini 3.8 Flash or GPT-6 Luna: which to pick

Pick Gemini 3.8 Flash for coding: it scores 61.6 on the Vibe Coding Index against 57.3 for GPT-6 Luna.

Pick GPT-6 Luna for output tokens 7.5x cheaper ($0.50 vs $3.75 per million) and a larger context window (1.05M against 1.05M tokens).

Model AGemini 3.8 FlashGoogle61.6/1006 independent benchmarks$2.74 per solved taskModel BGPT-6 LunaOpenAI57.3/1005 independent benchmarksNo cost per solved task
Shared results
102
Gemini 3.8 Flash only
26
GPT-6 Luna only
1
Like-for-like categories
8 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
Gemini 3.8 Flash 61.6 from 6 benchmarks, GPT-6 Luna 57.3 from 5 benchmarks.
Widest category
Reasoning, 16.9 points apart on average over 13 shared benchmarks.
Widest single gap
Mystery game puzzles: Gemini 3.8 Flash 47.0% against GPT-6 Luna 7.0%, 40.0 points apart on a 0 to 100 scale.
Output price
Gemini 3.8 Flash $3.75 against GPT-6 Luna $0.50 per million output tokens, 7.5x apart.
Context window
Gemini 3.8 Flash takes 1.05M tokens, GPT-6 Luna 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
CategoryGemini 3.8 FlashBarsGPT-6 LunaTies
Coding126
Intelligence20
Agentic work142
Reasoning130
Math231 tie
Knowledge70
Web design81
General283
All shared boards86151 tie
CodingLike-for-likeSix of these boards feed the Vibe Coding Index126
Coding benchmarks for Gemini 3.8 Flash and GPT-6 Luna
BenchmarkGemini 3.8 FlashGPT-6 LunaLeader
LiveBench coding82.2%84.5%GPT-6 Luna
LMArena WebDev15851578Gemini 3.8 Flash
AA Coding Index76.3Not measuredOnly Gemini 3.8 Flash measured
LiveCodeBench (Vals AI)89.5%Not measuredOnly Gemini 3.8 Flash measured
SWE-bench Verified (Vals AI)80.0%Not measuredOnly Gemini 3.8 Flash measured
Terminal-Bench 4 (Vals AI)19.7%13.6%Gemini 3.8 Flash
Vibe Code Bench (Vals AI)78.7%81.7%GPT-6 Luna
Code migration (Vals AI)36.5%42.5%GPT-6 Luna
ProgramBench (Vals AI)1.0%0.5%Gemini 3.8 Flash
IOI (Vals AI)56.9%55.6%Gemini 3.8 Flash
Terminal-Bench 2.1 (Vals AI)81.3%73.0%Gemini 3.8 Flash
SWE-bench Multimodal (Vals AI)40.8%37.5%Gemini 3.8 Flash
Vibe Code Bench 1-100 (Vals AI)18.8%Not measuredOnly Gemini 3.8 Flash measured
SkillsBench (Vals AI)58.0%Not measuredOnly Gemini 3.8 Flash measured
CyberBench (Vals AI)43.8%76.3%GPT-6 Luna
SciCode56.6%54.6%Gemini 3.8 Flash
FrontierCode41.2%42.4%GPT-6 Luna
FrontierSWE19.6%Not measuredOnly Gemini 3.8 Flash measured
WeirdML84.8%Not measuredOnly Gemini 3.8 Flash measured
ALE-Bench1269.71576.9GPT-6 Luna
LMArena Text: coding15081437Gemini 3.8 Flash
LMArena Text: software and IT15071423Gemini 3.8 Flash
LMArena WebDev: React15771573Gemini 3.8 Flash
LMArena WebDev: HTML16161595Gemini 3.8 Flash
LMArena WebDev: from an image15931571Gemini 3.8 Flash
SciCode12.3%Not measuredOnly Gemini 3.8 Flash measured
Neither model has a result on the other 13 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for Gemini 3.8 Flash and GPT-6 Luna
BenchmarkGemini 3.8 FlashGPT-6 LunaLeader
AA Intelligence Index40.938.1Gemini 3.8 Flash
Epoch Capabilities Index157.4156.8Gemini 3.8 Flash
Agentic workLike-for-likePublic boards, shown as published, kept out of the index142
Agentic work benchmarks for Gemini 3.8 Flash and GPT-6 Luna
BenchmarkGemini 3.8 FlashGPT-6 LunaLeader
AA Agentic Index40.2Not measuredOnly Gemini 3.8 Flash measured
Design Arena: agents building web apps12381191Gemini 3.8 Flash
Design Arena: agents building full-stack apps12441158Gemini 3.8 Flash
Design Arena: agents building mobile apps12361134Gemini 3.8 Flash
Design Arena: agents building Android apps1078Not measuredOnly Gemini 3.8 Flash measured
Design Arena: agents building Godot games1273Not measuredOnly Gemini 3.8 Flash measured
Design Arena: PowerPoint slides1153Not measuredOnly Gemini 3.8 Flash measured
Design Arena: spreadsheets1022Not measuredOnly Gemini 3.8 Flash measured
LiveBench research79.1%43.5%Gemini 3.8 Flash
LiveBench browser64.0%72.0%GPT-6 Luna
LiveBench long agentic30.1%48.1%GPT-6 Luna
SREBench (Vals AI)11.1%2.7%Gemini 3.8 Flash
Finance Agent (Vals AI)61.4%49.9%Gemini 3.8 Flash
Tax Agent Bench (Vals AI)32.1%19.9%Gemini 3.8 Flash
Legal Research Bench (Vals AI)38.9%30.3%Gemini 3.8 Flash
Harvey's Legal Agent Benchmark (Vals AI)12.1%4.2%Gemini 3.8 Flash
Excel Modeling Benchmark (Vals AI)72.2%68.5%Gemini 3.8 Flash
BioMysteryBench (Vals AI)62.2%61.5%Gemini 3.8 Flash
Terminal-Bench-Science (Vals AI)8.6%4.3%Gemini 3.8 Flash
Vals Science Index35.8%28.4%Gemini 3.8 Flash
Time Horizon Index (Vals AI)18.8%Not measuredOnly Gemini 3.8 Flash measured
CUA-bench (Vals AI)4.2%Not measuredOnly Gemini 3.8 Flash measured
APEX-Agents64.3%44.3%Gemini 3.8 Flash
Remote Labor Index5.8%Not measuredOnly Gemini 3.8 Flash measured
SpireClimbNot measured0.0%Only GPT-6 Luna measured
Vending-Bench 2$5,094Not measuredOnly Gemini 3.8 Flash measured
Neither model has a result on the other 15 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index130
Reasoning benchmarks for Gemini 3.8 Flash and GPT-6 Luna
BenchmarkGemini 3.8 FlashGPT-6 LunaLeader
GPQA Diamond94.4%Not measuredOnly Gemini 3.8 Flash measured
LiveBench reasoning89.3%81.8%Gemini 3.8 Flash
MysteryMechanism (Vals AI)36.5%19.4%Gemini 3.8 Flash
MMMU Pro (Vals AI)89.1%Not measuredOnly Gemini 3.8 Flash measured
GPQA Diamond (Artificial Analysis run)95.3%Not measuredOnly Gemini 3.8 Flash measured
Humanity's Last Exam (Artificial Analysis run)47.8%38.5%Gemini 3.8 Flash
GPQA Diamond (Epoch AI run)95.4%90.5%Gemini 3.8 Flash
ARC-AGI98.5%86.7%Gemini 3.8 Flash
ARC-AGI-289.2%59.3%Gemini 3.8 Flash
Chess puzzles58.0%36.0%Gemini 3.8 Flash
Mystery game puzzles47.0%7.0%Gemini 3.8 Flash
LMArena Text: hard prompts15091410Gemini 3.8 Flash
LMArena Text: expert15261442Gemini 3.8 Flash
LMArena Text: hard prompts in English15041410Gemini 3.8 Flash
LMArena Vision: diagrams13451245Gemini 3.8 Flash
LMArena Vision: homework13511261Gemini 3.8 Flash
Neither model has a result on the other 6 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index23
Math benchmarks for Gemini 3.8 Flash and GPT-6 Luna
BenchmarkGemini 3.8 FlashGPT-6 LunaLeader
OTIS Mock AIME98.9%98.9%Tie
FrontierMath, tiers 1 to 368.4%78.9%GPT-6 Luna
FrontierMath, tier 422.0%56.1%GPT-6 Luna
ProofBench48.0%64.0%GPT-6 Luna
LMArena Text: math15321409Gemini 3.8 Flash
LMArena Text: mathematical work15191427Gemini 3.8 Flash
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index70
Knowledge benchmarks for Gemini 3.8 Flash and GPT-6 Luna
BenchmarkGemini 3.8 FlashGPT-6 LunaLeader
MMLU Pro90.2%Not measuredOnly Gemini 3.8 Flash measured
LegalBench (Vals AI)87.0%Not measuredOnly Gemini 3.8 Flash measured
TaxEval (Vals AI)74.5%Not measuredOnly Gemini 3.8 Flash measured
MortgageTax (Vals AI)65.3%Not measuredOnly Gemini 3.8 Flash measured
MedCode (Vals AI)48.1%44.7%Gemini 3.8 Flash
SimpleQA Verified69.7%41.4%Gemini 3.8 Flash
LMArena Text: science15231386Gemini 3.8 Flash
LMArena Text: medicine and healthcare15081352Gemini 3.8 Flash
LMArena Text: legal and government14971400Gemini 3.8 Flash
LMArena Text: business and finance14791388Gemini 3.8 Flash
FACTS Benchmark Suite69.4%63.3%Gemini 3.8 Flash
Neither model has a result on the other 7 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index81
Web design benchmarks for Gemini 3.8 Flash and GPT-6 Luna
BenchmarkGemini 3.8 FlashGPT-6 LunaLeader
Design Arena: websites13001271Gemini 3.8 Flash
Design Arena: UI components13161295Gemini 3.8 Flash
Design Arena: game dev13041257Gemini 3.8 Flash
Design Arena: data viz12701216Gemini 3.8 Flash
Design Arena: 3D12941218Gemini 3.8 Flash
Design Arena: all code categories13031270Gemini 3.8 Flash
Design Arena: SVG12961252Gemini 3.8 Flash
Design Arena: ASCII art12621153Gemini 3.8 Flash
LiveBench design72.7%82.5%GPT-6 Luna
GeneralLike-for-likePublic boards, shown as published, kept out of the index283
General benchmarks for Gemini 3.8 Flash and GPT-6 Luna
BenchmarkGemini 3.8 FlashGPT-6 LunaLeader
Vals Index56.0%52.3%Gemini 3.8 Flash
LiveBench data analysis54.0%73.4%GPT-6 Luna
Public Benefits Bench (Vals AI)65.3%57.6%Gemini 3.8 Flash
SAGE (Vals AI)35.1%48.1%GPT-6 Luna
MedScribe (Vals AI)84.5%83.7%Gemini 3.8 Flash
Long context reasoning (AA-LCR)84.0%83.3%Gemini 3.8 Flash
DTBench95.7%90.1%Gemini 3.8 Flash
LMCA52.9%44.5%Gemini 3.8 Flash
Surface Evolver Bench76.9%Not measuredOnly Gemini 3.8 Flash measured
Furniture assembly31.7%44.2%GPT-6 Luna
LMArena Text14991389Gemini 3.8 Flash
LMArena Text: instruction following14921408Gemini 3.8 Flash
LMArena Text: creative writing14881363Gemini 3.8 Flash
LMArena Vision13151216Gemini 3.8 Flash
LMArena Text: English14981393Gemini 3.8 Flash
LMArena Text: non-English14911381Gemini 3.8 Flash
LMArena Text: Chinese15531427Gemini 3.8 Flash
LMArena Text: German15011376Gemini 3.8 Flash
LMArena Text: Japanese15071369Gemini 3.8 Flash
LMArena Text: Korean14641359Gemini 3.8 Flash
LMArena Text: French14961408Gemini 3.8 Flash
LMArena Text: Spanish14871391Gemini 3.8 Flash
LMArena Text: Russian15141392Gemini 3.8 Flash
LMArena Text: Polish14971383Gemini 3.8 Flash
LMArena Text: multi-turn15031392Gemini 3.8 Flash
LMArena Text: longer queries15071408Gemini 3.8 Flash
LMArena Text: writing and language14941384Gemini 3.8 Flash
LMArena Text: entertainment and media14801364Gemini 3.8 Flash
LMArena Vision: English13181204Gemini 3.8 Flash
LMArena Vision: Chinese14161266Gemini 3.8 Flash
LMArena Vision: reading text13331225Gemini 3.8 Flash
LMArena Vision: humor1321Not measuredOnly Gemini 3.8 Flash measured
LMArena Vision: creative writing12991205Gemini 3.8 Flash
Neither model has a result on the other 13 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Gemini 3.8 Flash and GPT-6 Luna
BenchmarkGemini 3.8 FlashGPT-6 LunaLeader
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

Gemini 3.8 Flash$0.0026Fits in one request
GPT-6 Luna$0.0003Fits in one request

GPT-6 Luna costs less

Repository review

50K input + 3K output tokens

Gemini 3.8 Flash$0.049Fits in one request
GPT-6 Luna$0.0065Fits in one request

GPT-6 Luna costs less

Agent loop

200K input + 10K output tokens

Gemini 3.8 Flash$0.188Fits in one request
GPT-6 Luna$0.025Fits in one request

GPT-6 Luna 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 Gemini 3.8 Flash and GPT-6 Luna
SpecificationGoogle Gemini 3.8 FlashOpenAI GPT-6 Luna
MakerGoogleOpenAI
WeightsNo public weights listedNo public weights listed
ReleasedSeptember 2, 2026September 22, 2026
Context window1.05M1.05M
Input $/Mtok$0.75$0.10
Output $/Mtok$3.75$0.50
Coding index61.657.3
Independent measurements behind it65
Cost per solved task$2.74Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Gemini 3.8 Flash better than GPT-6 Luna for coding?

Pick Gemini 3.8 Flash for coding: it scores 61.6 on the Vibe Coding Index against 57.3 for GPT-6 Luna. Pick GPT-6 Luna for output tokens 7.5x cheaper ($0.50 vs $3.75 per million) and a larger context window (1.05M against 1.05M tokens).

Which is cheaper, Gemini 3.8 Flash or GPT-6 Luna?

Gemini 3.8 Flash costs $0.75 per million input tokens and $3.75 per million output tokens; GPT-6 Luna costs $0.10 and $0.50. Cost per solved task is known for Gemini 3.8 Flash only.

Which has the bigger context window, Gemini 3.8 Flash or GPT-6 Luna?

GPT-6 Luna, with 1.05M tokens against 1.05M.

How many benchmarks is the Gemini 3.8 Flash vs GPT-6 Luna comparison based on?

102 benchmarks have a result for both models as of October 22, 2026. Gemini 3.8 Flash's coding index rests on 6 independent benchmarks and GPT-6 Luna's on 5. 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/gemini-3-8-flash-vs-gpt-6-luna" title="Gemini 3.8 Flash vs GPT-6 Luna, 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/gemini-3-8-flash-vs-gpt-6-luna">LLMBenchmark.ai, Gemini 3.8 Flash vs GPT-6 Luna</a></p>