Enter opens the highlighted result. Esc closes.

Gemini 3.5 Flash vs GPT-5.4

GPT-5.4 leads Gemini 3.5 Flash on coding, 64.4 to 39.8 on the Vibe Coding Index, and Gemini 3.5 Flash costs 1.7x less per output token ($9.00 vs $15.00 per million). Updated October 21, 2026; 95 benchmarks have a result for both.

Replace Gemini 3.5 Flash Replace GPT-5.4
Gemini 3.5 FlashGoogle
Coding
39.8
Overall
63.8
Output $/M
$9.00
56
GPT-5.4OpenAI
Coding
64.4
Overall
69.6
Output $/M
$15.00
39
vs boards won of 95 shared

Gemini 3.5 Flash or GPT-5.4: which to pick

Pick GPT-5.4 for coding: it scores 64.4 on the Vibe Coding Index against 39.8 for Gemini 3.5 Flash, though on 3 benchmarks against 4.

Pick Gemini 3.5 Flash for output tokens 1.7x cheaper ($9.00 vs $15.00 per million) and work measured by LMArena WebDev or LiveCodeBench (Vals AI), where it leads.

Model AGemini 3.5 FlashGoogle39.8/1004 independent benchmarks$1.21 per solved taskModel BGPT-5.4OpenAI64.4/1003 independent benchmarks$1.02 per solved task
Shared results
95
Gemini 3.5 Flash only
17
GPT-5.4 only
15
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 3.5 Flash 39.8 from 4 benchmarks, GPT-5.4 64.4 from 3 benchmarks.
Widest category
Agentic work, 14.1 points apart on average over 8 shared benchmarks.
Widest single gap
ProofBench: Gemini 3.5 Flash 31.0% against GPT-5.4 56.0%, 25.0 points apart on a 0 to 100 scale.
Output price
Gemini 3.5 Flash $9.00 against GPT-5.4 $15.00 per million output tokens, 1.7x apart.
Context window
Gemini 3.5 Flash takes 1.05M tokens, GPT-5.4 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 3.5 FlashBarsGPT-5.4Ties
Coding87
Intelligence02
Agentic work53
Reasoning510
Math15
Knowledge74
Web design80
General228
All shared boards5639
CodingLike-for-likeSix of these boards feed the Vibe Coding Index87
Coding benchmarks for Gemini 3.5 Flash and GPT-5.4
BenchmarkGemini 3.5 FlashGPT-5.4Leader
LMArena WebDev14991464Gemini 3.5 Flash
AA Coding IndexNot measured71.1Only GPT-5.4 measured
LiveCodeBench (Vals AI)87.6%84.1%Gemini 3.5 Flash
SWE-bench Verified (Vals AI)78.8%78.2%Gemini 3.5 Flash
Terminal-Bench 4 (Vals AI)6.1%Not measuredOnly Gemini 3.5 Flash measured
Vibe Code Bench (Vals AI)48.7%67.4%GPT-5.4
Code migration (Vals AI)26.8%35.0%GPT-5.4
ProgramBench (Vals AI)0.0%0.5%GPT-5.4
Terminal-Bench 2.1 (Vals AI)74.2%Not measuredOnly Gemini 3.5 Flash measured
SkillsBench (Vals AI)52.7%51.7%Gemini 3.5 Flash
SciCode53.9%Not measuredOnly Gemini 3.5 Flash measured
Terminal-Bench Hard (Artificial Analysis run)46.2%57.6%GPT-5.4
Terminal-BenchNot measured81.8%Only GPT-5.4 measured
GSO-BenchNot measured31.4%Only GPT-5.4 measured
WeirdML62.6%77.7%GPT-5.4
MirrorCodeNot measured15.6%Only GPT-5.4 measured
ALE-Bench911.01607.0GPT-5.4
LMArena Text: coding14931496GPT-5.4
LMArena Text: software and IT14941491Gemini 3.5 Flash
LMArena WebDev: React14951454Gemini 3.5 Flash
LMArena WebDev: HTML15221478Gemini 3.5 Flash
LMArena WebDev: from an image15041436Gemini 3.5 Flash
SciCode12.3%Not measuredOnly Gemini 3.5 Flash measured
Neither model has a result on the other 16 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for Gemini 3.5 Flash and GPT-5.4
BenchmarkGemini 3.5 FlashGPT-5.4Leader
AA Intelligence Index33.639.0GPT-5.4
Epoch Capabilities Index154.9157.3GPT-5.4
Agentic workLike-for-likePublic boards, shown as published, kept out of the index53
Agentic work benchmarks for Gemini 3.5 Flash and GPT-5.4
BenchmarkGemini 3.5 FlashGPT-5.4Leader
Design Arena: agents building web apps12021059Gemini 3.5 Flash
Design Arena: agents building full-stack apps11971017Gemini 3.5 Flash
Design Arena: agents building mobile apps11741087Gemini 3.5 Flash
Design Arena: agents building Android apps11841030Gemini 3.5 Flash
Design Arena: agents building Godot games10981135GPT-5.4
Design Arena: agents building games1180Not measuredOnly Gemini 3.5 Flash measured
Design Arena: HTML slides1142Not measuredOnly Gemini 3.5 Flash measured
Design Arena: PowerPoint slides1179Not measuredOnly Gemini 3.5 Flash measured
Finance Agent (Vals AI)57.9%Not measuredOnly Gemini 3.5 Flash measured
Tax Agent Bench (Vals AI)21.6%Not measuredOnly Gemini 3.5 Flash measured
Legal Research Bench (Vals AI)30.8%Not measuredOnly Gemini 3.5 Flash measured
Excel Modeling Benchmark (Vals AI)63.5%Not measuredOnly Gemini 3.5 Flash measured
Terminal-Bench-Science (Vals AI)5.7%Not measuredOnly Gemini 3.5 Flash measured
τ²-Bench Telecom95.6%87.1%Gemini 3.5 Flash
APEX-Agents27.5%52.4%GPT-5.4
METR time horizonsNot measured74.3%Only GPT-5.4 measured
DeepResearch BenchNot measured35.1%Only GPT-5.4 measured
PostTrainBenchNot measured19.0%Only GPT-5.4 measured
Vending-Bench 2$5,396$6,144GPT-5.4
Neither model has a result on the other 22 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index510
Reasoning benchmarks for Gemini 3.5 Flash and GPT-5.4
BenchmarkGemini 3.5 FlashGPT-5.4Leader
GPQA Diamond92.7%91.7%Gemini 3.5 Flash
MMMU Pro (Vals AI)88.3%87.5%Gemini 3.5 Flash
GPQA Diamond (Artificial Analysis run)92.2%92.0%Gemini 3.5 Flash
Humanity's Last Exam (Artificial Analysis run)42.7%43.7%GPT-5.4
GPQA Diamond (Epoch AI run)92.8%93.3%GPT-5.4
ARC-AGI92.5%93.7%GPT-5.4
ARC-AGI-272.1%74.0%GPT-5.4
SimpleBench76.7%Not measuredOnly Gemini 3.5 Flash measured
Chess puzzles45.0%48.0%GPT-5.4
Mystery game puzzles32.0%37.0%GPT-5.4
EBR-bench4.8%25.4%GPT-5.4
LMArena Text: hard prompts14881485Gemini 3.5 Flash
LMArena Text: expert14951507GPT-5.4
LMArena Text: hard prompts in English14831479Gemini 3.5 Flash
LMArena Vision: diagrams13231327GPT-5.4
LMArena Vision: homework13301336GPT-5.4
Kagi LLM BenchmarkNot measured63.8%Only GPT-5.4 measured
Neither model has a result on the other 5 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index15
Math benchmarks for Gemini 3.5 Flash and GPT-5.4
BenchmarkGemini 3.5 FlashGPT-5.4Leader
AIMENot measured96.7%Only GPT-5.4 measured
OTIS Mock AIME95.6%97.8%GPT-5.4
FrontierMath, tiers 1 to 362.8%78.6%GPT-5.4
FrontierMath, tier 426.8%49.0%GPT-5.4
ProofBench31.0%56.0%GPT-5.4
LMArena Text: math15041488Gemini 3.5 Flash
LMArena Text: mathematical work14951497GPT-5.4
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index74
Knowledge benchmarks for Gemini 3.5 Flash and GPT-5.4
BenchmarkGemini 3.5 FlashGPT-5.4Leader
MMLU Pro89.5%87.5%Gemini 3.5 Flash
LegalBench (Vals AI)83.6%86.0%GPT-5.4
CaseLaw (Vals AI)Not measured63.8%Only GPT-5.4 measured
CorpFin (Vals AI)64.7%65.3%GPT-5.4
TaxEval (Vals AI)74.4%74.0%Gemini 3.5 Flash
MortgageTax (Vals AI)68.1%68.3%GPT-5.4
MedQA (Vals AI)Not measured96.1%Only GPT-5.4 measured
MedCode (Vals AI)55.8%41.3%Gemini 3.5 Flash
SimpleQA Verified66.2%45.1%Gemini 3.5 Flash
LMArena Text: science14951479Gemini 3.5 Flash
LMArena Text: medicine and healthcare14831458Gemini 3.5 Flash
LMArena Text: legal and government14881486Gemini 3.5 Flash
LMArena Text: business and finance14671474GPT-5.4
FACTS Benchmark Suite63.1%Not measuredOnly Gemini 3.5 Flash measured
Neither model has a result on the other 4 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index80
Web design benchmarks for Gemini 3.5 Flash and GPT-5.4
BenchmarkGemini 3.5 FlashGPT-5.4Leader
Design Arena: websites12681199Gemini 3.5 Flash
Design Arena: UI components12731215Gemini 3.5 Flash
Design Arena: game dev12631195Gemini 3.5 Flash
Design Arena: data viz12361196Gemini 3.5 Flash
Design Arena: 3D12441066Gemini 3.5 Flash
Design Arena: all code categories12661180Gemini 3.5 Flash
Design Arena: SVG12541142Gemini 3.5 Flash
Design Arena: ASCII art12541188Gemini 3.5 Flash
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index228
General benchmarks for Gemini 3.5 Flash and GPT-5.4
BenchmarkGemini 3.5 FlashGPT-5.4Leader
Public Benefits Bench (Vals AI)59.5%Not measuredOnly Gemini 3.5 Flash measured
SAGE (Vals AI)49.9%43.3%Gemini 3.5 Flash
MedScribe (Vals AI)76.6%77.5%GPT-5.4
Vals Multimodal Index62.9%Not measuredOnly Gemini 3.5 Flash measured
IFBench76.3%73.9%Gemini 3.5 Flash
Long context reasoning (AA-LCR)74.3%82.0%GPT-5.4
DTBench94.7%94.4%Gemini 3.5 Flash
LMCA47.1%52.0%GPT-5.4
CL-benchNot measured27.9%Only GPT-5.4 measured
Surface Evolver Bench58.1%Not measuredOnly Gemini 3.5 Flash measured
Furniture assemblyNot measured37.5%Only GPT-5.4 measured
CL-bench LifeNot measured21.7%Only GPT-5.4 measured
LMArena Text14821469Gemini 3.5 Flash
LMArena Text: instruction following14671469GPT-5.4
LMArena Text: creative writing14701439Gemini 3.5 Flash
LMArena Vision13081303Gemini 3.5 Flash
LMArena Text: English14811467Gemini 3.5 Flash
LMArena Text: non-English14751465Gemini 3.5 Flash
LMArena Text: Chinese15271520Gemini 3.5 Flash
LMArena Text: German14941472Gemini 3.5 Flash
LMArena Text: Japanese14861485Gemini 3.5 Flash
LMArena Text: Korean14481449GPT-5.4
LMArena Text: French14901493GPT-5.4
LMArena Text: Spanish14801452Gemini 3.5 Flash
LMArena Text: Russian14931480Gemini 3.5 Flash
LMArena Text: Polish15161472Gemini 3.5 Flash
LMArena Text: multi-turn14811482GPT-5.4
LMArena Text: longer queries14831473Gemini 3.5 Flash
LMArena Text: writing and language14781458Gemini 3.5 Flash
LMArena Text: entertainment and media14591438Gemini 3.5 Flash
LMArena Vision: English13051301Gemini 3.5 Flash
LMArena Vision: Chinese13771354Gemini 3.5 Flash
LMArena Vision: reading text13171312Gemini 3.5 Flash
LMArena Vision: humor13211284Gemini 3.5 Flash
LMArena Vision: recognizing thingsNot measured1251Only GPT-5.4 measured
LMArena Vision: creative writing13011283Gemini 3.5 Flash
LMArena Document14631471GPT-5.4
Neither model has a result on the other 9 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Gemini 3.5 Flash and GPT-5.4
BenchmarkGemini 3.5 FlashGPT-5.4Leader
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 3.5 Flash$0.0060Fits in one request
GPT-5.4$0.010Fits in one request

Gemini 3.5 Flash costs less

Repository review

50K input + 3K output tokens

Gemini 3.5 Flash$0.102Fits in one request
GPT-5.4$0.170Fits in one request

Gemini 3.5 Flash costs less

Agent loop

200K input + 10K output tokens

Gemini 3.5 Flash$0.390Fits in one request
GPT-5.4$0.650Fits in one request

Gemini 3.5 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 3.5 Flash and GPT-5.4
SpecificationGoogle Gemini 3.5 FlashOpenAI GPT-5.4
MakerGoogleOpenAI
WeightsNo public weights listedNo public weights listed
ReleasedMay 19, 2026March 5, 2026
Context window1.05M1.05M
Input $/Mtok$1.50$2.50
Output $/Mtok$9.00$15.00
Coding index39.864.4
Independent measurements behind it43
Cost per solved task$1.21$1.02
SourceOpenRouter listingOpenRouter listing

Questions

Is Gemini 3.5 Flash better than GPT-5.4 for coding?

Pick GPT-5.4 for coding: it scores 64.4 on the Vibe Coding Index against 39.8 for Gemini 3.5 Flash, though on 3 benchmarks against 4. Pick Gemini 3.5 Flash for output tokens 1.7x cheaper ($9.00 vs $15.00 per million) and work measured by LMArena WebDev or LiveCodeBench (Vals AI), where it leads.

Which is cheaper, Gemini 3.5 Flash or GPT-5.4?

Gemini 3.5 Flash costs $1.50 per million input tokens and $9.00 per million output tokens; GPT-5.4 costs $2.50 and $15.00. Per solved SWE-bench Verified task, Gemini 3.5 Flash costs $1.21 and GPT-5.4 $1.02.

Which has the bigger context window, Gemini 3.5 Flash or GPT-5.4?

GPT-5.4, with 1.05M tokens against 1.05M.

How many benchmarks is the Gemini 3.5 Flash vs GPT-5.4 comparison based on?

95 benchmarks have a result for both models as of October 21, 2026. Gemini 3.5 Flash's coding index rests on 4 independent benchmarks and GPT-5.4's on 3. 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-3-5-flash-vs-gpt-5-4" title="Gemini 3.5 Flash vs GPT-5.4, 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-3-5-flash-vs-gpt-5-4">LLMBenchmark.ai, Gemini 3.5 Flash vs GPT-5.4</a></p>