LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

Gemini 3.8 Flash vs GPT-5.6 Sol

Updated October 21, 2026. 120 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace Gemini 3.8 Flash Replace GPT-5.6 Sol
Gemini 3.8 FlashGoogle
Coding
70.5
Overall
40.9
Output $/M
$4.125
50
GPT-5.6 SolOpenAI
Coding
76.7
Overall
47.0
Output $/M
$20.00
68
vs boards won of 120 shared, 2 tied

Decision reading

Pick GPT-5.6 Sol for coding: it scores 76.7 on the Vibe Coding Index against 70.5 for Gemini 3.8 Flash.

Pick Gemini 3.8 Flash when output tokens cost 4.8x less, or when what matters is LiveCodeBench (Vals AI) or SkillsBench (Vals AI), where it leads.

Model AGemini 3.8 FlashGoogle70.5/1002 independent benchmarks$2.74 per solved taskModel BGPT-5.6 SolOpenAI76.7/1002 independent benchmarks$1.20 per solved task
Shared results
120
Gemini 3.8 Flash only
6
GPT-5.6 Sol only
16
Like-for-like categories
8 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
Gemini 3.8 Flash 70.5 from 2 benchmarks, GPT-5.6 Sol 76.7 from 2 benchmarks.
Widest category
Math, 21.9 points apart on average over 6 shared benchmarks.
Widest single gap
FrontierMath, tier 4: Gemini 3.8 Flash 22.0% against GPT-5.6 Sol 82.9%, 61.0 points apart on a 0 to 100 scale.
Output price
Gemini 3.8 Flash $4.125 against GPT-5.6 Sol $20.00 per million output tokens, 4.8x apart.
Context window
Gemini 3.8 Flash takes 1.05M tokens, GPT-5.6 Sol 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-5.6 SolTies
Coding420
Intelligence02
Agentic work417
Reasoning106
Math24
Knowledge721 tie
Web design09
General2381 tie
All shared boards50682 ties
CodingLike-for-likeFour of these boards feed the Vibe Coding Index420
Coding benchmarks for Gemini 3.8 Flash and GPT-5.6 Sol
BenchmarkGemini 3.8 FlashGPT-5.6 SolLeader
LiveBench coding82.2%89.7%GPT-5.6 Sol
LMArena WebDev15851616GPT-5.6 Sol
AA Coding Index76.377.4GPT-5.6 Sol
LiveCodeBench (Vals AI)89.5%82.6%Gemini 3.8 Flash
SWE-bench Verified (Vals AI)80.0%96.2%GPT-5.6 Sol
Terminal-Bench 4 (Vals AI)19.7%37.9%GPT-5.6 Sol
Vibe Code Bench (Vals AI)78.7%80.5%GPT-5.6 Sol
Code migration (Vals AI)36.5%52.9%GPT-5.6 Sol
ProgramBench (Vals AI)1.0%1.5%GPT-5.6 Sol
IOI (Vals AI)56.9%91.2%GPT-5.6 Sol
Terminal-Bench 2.1 (Vals AI)81.3%85.8%GPT-5.6 Sol
SWE-bench Multimodal (Vals AI)40.8%Not measuredOnly Gemini 3.8 Flash measured
Vibe Code Bench 1-100 (Vals AI)18.8%20.0%GPT-5.6 Sol
SkillsBench (Vals AI)58.0%54.1%Gemini 3.8 Flash
CyberBench (Vals AI)43.8%76.3%GPT-5.6 Sol
SciCode56.6%57.8%GPT-5.6 Sol
Terminal-Bench Hard (Artificial Analysis run)Not measured65.9%Only GPT-5.6 Sol measured
FrontierCode41.2%47.5%GPT-5.6 Sol
FrontierSWE19.6%32.2%GPT-5.6 Sol
GSO-BenchNot measured76.5%Only GPT-5.6 Sol measured
WeirdML84.8%88.8%GPT-5.6 Sol
MirrorCodeNot measured20.0%Only GPT-5.6 Sol measured
ALE-Bench1269.72176.9GPT-5.6 Sol
LMArena Text: coding15081496Gemini 3.8 Flash
LMArena Text: software and IT15071489Gemini 3.8 Flash
LMArena WebDev: React15771620GPT-5.6 Sol
LMArena WebDev: HTML16161618GPT-5.6 Sol
LMArena WebDev: from an image15931609GPT-5.6 Sol
Neither model has a result on the other 10 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for Gemini 3.8 Flash and GPT-5.6 Sol
BenchmarkGemini 3.8 FlashGPT-5.6 SolLeader
AA Intelligence Index40.947.0GPT-5.6 Sol
Epoch Capabilities Index157.4162.8GPT-5.6 Sol
Agentic workLike-for-likePublic boards, shown as published, kept out of the index417
Agentic work benchmarks for Gemini 3.8 Flash and GPT-5.6 Sol
BenchmarkGemini 3.8 FlashGPT-5.6 SolLeader
AA Agentic Index40.250.2GPT-5.6 Sol
Design Arena: agents building web apps12381245GPT-5.6 Sol
Design Arena: agents building full-stack apps12441261GPT-5.6 Sol
Design Arena: agents building mobile apps12361249GPT-5.6 Sol
Design Arena: agents building Android apps10781246GPT-5.6 Sol
Design Arena: agents building Godot games12731260Gemini 3.8 Flash
Design Arena: agents building gamesNot measured1267Only GPT-5.6 Sol measured
Design Arena: HTML slidesNot measured1286Only GPT-5.6 Sol measured
Design Arena: PowerPoint slides11781240GPT-5.6 Sol
Design Arena: spreadsheets1052Not measuredOnly Gemini 3.8 Flash measured
LiveBench research79.1%54.5%Gemini 3.8 Flash
LiveBench browser64.0%88.0%GPT-5.6 Sol
LiveBench long agentic30.1%48.5%GPT-5.6 Sol
SREBench (Vals AI)11.1%30.5%GPT-5.6 Sol
Finance Agent (Vals AI)61.4%53.8%Gemini 3.8 Flash
Tax Agent Bench (Vals AI)32.1%31.1%Gemini 3.8 Flash
Legal Research Bench (Vals AI)38.9%48.1%GPT-5.6 Sol
Harvey's Legal Agent Benchmark (Vals AI)12.1%Not measuredOnly Gemini 3.8 Flash measured
Excel Modeling Benchmark (Vals AI)72.2%72.3%GPT-5.6 Sol
BioMysteryBench (Vals AI)62.2%71.1%GPT-5.6 Sol
Terminal-Bench-Science (Vals AI)8.6%20.0%GPT-5.6 Sol
Vals Science Index35.8%41.5%GPT-5.6 Sol
Time Horizon Index (Vals AI)18.8%23.8%GPT-5.6 Sol
CUA-bench (Vals AI)4.2%8.3%GPT-5.6 Sol
τ²-Bench TelecomNot measured85.1%Only GPT-5.6 Sol measured
APEX-Agents64.3%Not measuredOnly Gemini 3.8 Flash measured
Remote Labor Index5.8%Not measuredOnly Gemini 3.8 Flash measured
OSWorld 2.0Not measured27.3%Only GPT-5.6 Sol measured
BALROGNot measured60.0%Only GPT-5.6 Sol measured
PostTrainBenchNot measured36.2%Only GPT-5.6 Sol measured
Vending-Bench 2$5,094$9,619GPT-5.6 Sol
LMArena SearchNot measured1261Only GPT-5.6 Sol measured
Neither model has a result on the other 7 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index106
Reasoning benchmarks for Gemini 3.8 Flash and GPT-5.6 Sol
BenchmarkGemini 3.8 FlashGPT-5.6 SolLeader
GPQA Diamond94.4%95.2%GPT-5.6 Sol
LiveBench reasoning89.3%91.7%GPT-5.6 Sol
MysteryMechanism (Vals AI)36.5%33.3%Gemini 3.8 Flash
MMMU Pro (Vals AI)89.1%88.8%Gemini 3.8 Flash
GPQA Diamond (Artificial Analysis run)95.3%94.1%Gemini 3.8 Flash
Humanity's Last Exam (Artificial Analysis run)47.8%49.5%GPT-5.6 Sol
GPQA Diamond (Epoch AI run)95.4%93.5%Gemini 3.8 Flash
ARC-AGI98.5%97.5%Gemini 3.8 Flash
ARC-AGI-289.2%92.5%GPT-5.6 Sol
SimpleBenchNot measured64.8%Only GPT-5.6 Sol measured
Chess puzzles58.0%59.0%GPT-5.6 Sol
Mystery game puzzles47.0%58.0%GPT-5.6 Sol
EBR-benchNot measured44.8%Only GPT-5.6 Sol measured
LMArena Text: hard prompts15091483Gemini 3.8 Flash
LMArena Text: expert15261517Gemini 3.8 Flash
LMArena Text: hard prompts in English15041477Gemini 3.8 Flash
LMArena Vision: diagrams13451305Gemini 3.8 Flash
LMArena Vision: homework13511305Gemini 3.8 Flash
Neither model has a result on the other 3 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index24
Math benchmarks for Gemini 3.8 Flash and GPT-5.6 Sol
BenchmarkGemini 3.8 FlashGPT-5.6 SolLeader
OTIS Mock AIME98.9%100.0%GPT-5.6 Sol
FrontierMath, tiers 1 to 368.4%89.1%GPT-5.6 Sol
FrontierMath, tier 422.0%82.9%GPT-5.6 Sol
ProofBench48.0%83.0%GPT-5.6 Sol
LMArena Text: math15321473Gemini 3.8 Flash
LMArena Text: mathematical work15191496Gemini 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 index72
Knowledge benchmarks for Gemini 3.8 Flash and GPT-5.6 Sol
BenchmarkGemini 3.8 FlashGPT-5.6 SolLeader
MMLU Pro90.2%89.1%Gemini 3.8 Flash
LegalBench (Vals AI)87.0%87.0%Gemini 3.8 Flash
CorpFin (Vals AI)Not measured64.4%Only GPT-5.6 Sol measured
TaxEval (Vals AI)74.5%74.8%GPT-5.6 Sol
MortgageTax (Vals AI)65.3%67.3%GPT-5.6 Sol
MedCode (Vals AI)48.1%44.0%Gemini 3.8 Flash
SimpleQA Verified69.7%69.7%Tie
LMArena Text: science15231455Gemini 3.8 Flash
LMArena Text: medicine and healthcare15081431Gemini 3.8 Flash
LMArena Text: legal and government14971476Gemini 3.8 Flash
LMArena Text: business and finance14791457Gemini 3.8 Flash
Neither model has a result on the other 6 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index09
Web design benchmarks for Gemini 3.8 Flash and GPT-5.6 Sol
BenchmarkGemini 3.8 FlashGPT-5.6 SolLeader
Design Arena: websites13011321GPT-5.6 Sol
Design Arena: UI components13181338GPT-5.6 Sol
Design Arena: game dev13051368GPT-5.6 Sol
Design Arena: data viz12691323GPT-5.6 Sol
Design Arena: 3D12951385GPT-5.6 Sol
Design Arena: all code categories13041341GPT-5.6 Sol
Design Arena: SVG12951316GPT-5.6 Sol
Design Arena: ASCII art12641302GPT-5.6 Sol
LiveBench design72.7%81.6%GPT-5.6 Sol
GeneralLike-for-likePublic boards, shown as published, kept out of the index238
General benchmarks for Gemini 3.8 Flash and GPT-5.6 Sol
BenchmarkGemini 3.8 FlashGPT-5.6 SolLeader
Vals Index56.0%Not measuredOnly Gemini 3.8 Flash measured
LiveBench data analysis54.0%79.8%GPT-5.6 Sol
Public Benefits Bench (Vals AI)65.3%66.5%GPT-5.6 Sol
SAGE (Vals AI)35.1%52.6%GPT-5.6 Sol
MedScribe (Vals AI)84.5%85.2%GPT-5.6 Sol
Vals Multimodal IndexNot measured72.6%Only GPT-5.6 Sol measured
IFBenchNot measured72.7%Only GPT-5.6 Sol measured
Long context reasoning (AA-LCR)84.0%84.0%Tie
DTBench95.7%95.7%GPT-5.6 Sol
LMCA52.9%58.5%GPT-5.6 Sol
Surface Evolver Bench76.9%93.1%GPT-5.6 Sol
Furniture assembly31.7%56.7%GPT-5.6 Sol
LMArena Text14991457Gemini 3.8 Flash
LMArena Text: instruction following14921483Gemini 3.8 Flash
LMArena Text: creative writing14881449Gemini 3.8 Flash
LMArena Vision13151281Gemini 3.8 Flash
LMArena Text: English14981457Gemini 3.8 Flash
LMArena Text: non-English14911451Gemini 3.8 Flash
LMArena Text: Chinese15531525Gemini 3.8 Flash
LMArena Text: German15011476Gemini 3.8 Flash
LMArena Text: Japanese15071472Gemini 3.8 Flash
LMArena Text: Korean14641442Gemini 3.8 Flash
LMArena Text: French14961472Gemini 3.8 Flash
LMArena Text: Spanish14871443Gemini 3.8 Flash
LMArena Text: Russian15141471Gemini 3.8 Flash
LMArena Text: Polish14971459Gemini 3.8 Flash
LMArena Text: multi-turn15031461Gemini 3.8 Flash
LMArena Text: longer queries15071481Gemini 3.8 Flash
LMArena Text: writing and language14941462Gemini 3.8 Flash
LMArena Text: entertainment and media14801450Gemini 3.8 Flash
LMArena Vision: English13181280Gemini 3.8 Flash
LMArena Vision: Chinese14161313Gemini 3.8 Flash
LMArena Vision: reading text13331290Gemini 3.8 Flash
LMArena Vision: humor13211246Gemini 3.8 Flash
LMArena Vision: creative writing12991283Gemini 3.8 Flash
LMArena DocumentNot measured1483Only GPT-5.6 Sol measured
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 Gemini 3.8 Flash and GPT-5.6 Sol
BenchmarkGemini 3.8 FlashGPT-5.6 SolLeader
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$0.0029Fits in one request
GPT-5.6 Sol$0.014Fits in one request

Gemini 3.8 Flash costs less

Repository review

50K input + 3K output tokens

Gemini 3.8 Flash$0.054Fits in one request
GPT-5.6 Sol$0.260Fits in one request

Gemini 3.8 Flash costs less

Agent loop

200K input + 10K output tokens

Gemini 3.8 Flash$0.206Fits in one request
GPT-5.6 Sol$1.00Fits in one request

Gemini 3.8 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 Gemini 3.8 Flash and GPT-5.6 Sol
SpecificationGoogle Gemini 3.8 FlashOpenAI GPT-5.6 Sol
MakerGoogleOpenAI
WeightsNo public weights listedNo public weights listed
ReleasedSeptember 2, 2026July 9, 2026
Context window1.05M1.05M
Input $/Mtok$0.825$4.00
Output $/Mtok$4.125$20.00
Coding index70.576.7
Independent measurements behind it22
Cost per solved task$2.74$1.20
SourceOpenRouter listingOpenRouter listing

Questions

Is Gemini 3.8 Flash better than GPT-5.6 Sol for coding?

Pick GPT-5.6 Sol for coding: it scores 76.7 on the Vibe Coding Index against 70.5 for Gemini 3.8 Flash. Pick Gemini 3.8 Flash when output tokens cost 4.8x less, or when what matters is LiveCodeBench (Vals AI) or SkillsBench (Vals AI), where it leads.

Which is cheaper, Gemini 3.8 Flash or GPT-5.6 Sol?

Gemini 3.8 Flash costs $0.825 per million input tokens and $4.125 per million output tokens; GPT-5.6 Sol costs $4.00 and $20.00. Per solved SWE-bench Verified task, Gemini 3.8 Flash costs $2.74 and GPT-5.6 Sol $1.20.

Which has the bigger context window, Gemini 3.8 Flash or GPT-5.6 Sol?

GPT-5.6 Sol, with 1.05M tokens against 1.05M.

How many benchmarks is the Gemini 3.8 Flash vs GPT-5.6 Sol comparison based on?

120 benchmarks have a result for both models as of October 21, 2026. Gemini 3.8 Flash's coding index rests on 2 independent benchmarks and GPT-5.6 Sol'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/gemini-3-8-flash-vs-gpt-5-6-sol" title="Gemini 3.8 Flash vs GPT-5.6 Sol, 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-5-6-sol">LLMBenchmark.ai, Gemini 3.8 Flash vs GPT-5.6 Sol</a></p>