Enter opens the highlighted result. Esc closes.

GPT-5.4 vs GPT-5.5

GPT-5.5 leads GPT-5.4 on coding, 71.3 to 64.4 on the Vibe Coding Index, and GPT-5.4 costs 2.0x less per output token ($15.00 vs $30.00 per million). Updated October 21, 2026; 106 benchmarks have a result for both.

Replace GPT-5.4 Replace GPT-5.5
GPT-5.4OpenAI
Coding
64.4
Overall
69.6
Output $/M
$15.00
18
GPT-5.5OpenAI
Coding
71.3
Overall
69.2
Output $/M
$30.00
87
vs boards won of 106 shared, 1 tied

GPT-5.4 or GPT-5.5: which to pick

Pick GPT-5.5 for coding: it scores 71.3 on the Vibe Coding Index against 64.4 for GPT-5.4.

Pick GPT-5.4 for output tokens 2.0x cheaper ($15.00 vs $30.00 per million) and a lower cost per solved task ($1.02 against $1.65).

Model AGPT-5.4OpenAI64.4/1003 independent benchmarks$1.02 per solved taskModel BGPT-5.5OpenAI71.3/1005 independent benchmarks$1.65 per solved task
Shared results
106
GPT-5.4 only
4
GPT-5.5 only
28
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
GPT-5.4 64.4 from 3 benchmarks, GPT-5.5 71.3 from 5 benchmarks.
Widest category
Web design, 10.0 points apart on average over 8 shared benchmarks.
Widest single gap
Kagi LLM Benchmark: GPT-5.4 63.8% against GPT-5.5 88.8%, 25.0 points apart on a 0 to 100 scale.
Output price
GPT-5.4 $15.00 against GPT-5.5 $30.00 per million output tokens, 2.0x apart.
Context window
Both take 1.05M tokens.

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
CategoryGPT-5.4BarsGPT-5.5Ties
Coding2161 tie
Intelligence11
Agentic work010
Reasoning313
Math42
Knowledge012
Web design08
General825
All shared boards18871 tie
CodingLike-for-likeSix of these boards feed the Vibe Coding Index216
Coding benchmarks for GPT-5.4 and GPT-5.5
BenchmarkGPT-5.4GPT-5.5Leader
LiveBench codingNot measured85.1%Only GPT-5.5 measured
LMArena WebDev14641513GPT-5.5
AA Coding Index71.174.9GPT-5.5
LiveCodeBench (Vals AI)84.1%85.3%GPT-5.5
SWE-bench Verified (Vals AI)78.2%82.6%GPT-5.5
Vibe Code Bench (Vals AI)67.4%69.8%GPT-5.5
Code migration (Vals AI)35.0%45.2%GPT-5.5
ProgramBench (Vals AI)0.5%0.5%Tie
Terminal-Bench 2.1 (Vals AI)Not measured76.4%Only GPT-5.5 measured
SkillsBench (Vals AI)51.7%62.5%GPT-5.5
SciCodeNot measured56.1%Only GPT-5.5 measured
Terminal-Bench Hard (Artificial Analysis run)57.6%60.6%GPT-5.5
Terminal-Bench81.8%84.7%GPT-5.5
FrontierCodeNot measured43.0%Only GPT-5.5 measured
GSO-Bench31.4%40.2%GPT-5.5
WeirdML77.7%84.9%GPT-5.5
MirrorCode15.6%10.0%GPT-5.4
ALE-Bench1607.01943.0GPT-5.5
LMArena Text: coding14961494GPT-5.4
LMArena Text: software and IT14911493GPT-5.5
LMArena WebDev: React14541506GPT-5.5
LMArena WebDev: HTML14781543GPT-5.5
LMArena WebDev: from an image14361532GPT-5.5
Neither model has a result on the other 16 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11
Intelligence benchmarks for GPT-5.4 and GPT-5.5
BenchmarkGPT-5.4GPT-5.5Leader
AA Intelligence Index39.038.4GPT-5.4
Epoch Capabilities Index157.3159.7GPT-5.5
Agentic workLike-for-likePublic boards, shown as published, kept out of the index010
Agentic work benchmarks for GPT-5.4 and GPT-5.5
BenchmarkGPT-5.4GPT-5.5Leader
AA Agentic IndexNot measured36.4Only GPT-5.5 measured
Design Arena: agents building web apps10591117GPT-5.5
Design Arena: agents building full-stack apps10171089GPT-5.5
Design Arena: agents building mobile apps10871143GPT-5.5
Design Arena: agents building Android apps10301163GPT-5.5
Design Arena: agents building Godot games11351173GPT-5.5
Design Arena: agents building gamesNot measured1179Only GPT-5.5 measured
Design Arena: HTML slidesNot measured1067Only GPT-5.5 measured
Design Arena: PowerPoint slidesNot measured1109Only GPT-5.5 measured
LiveBench researchNot measured61.6%Only GPT-5.5 measured
LiveBench browserNot measured48.0%Only GPT-5.5 measured
LiveBench long agenticNot measured44.6%Only GPT-5.5 measured
SREBench (Vals AI)Not measured3.8%Only GPT-5.5 measured
Finance Agent (Vals AI)Not measured51.8%Only GPT-5.5 measured
Tax Agent Bench (Vals AI)Not measured21.8%Only GPT-5.5 measured
Legal Research Bench (Vals AI)Not measured40.4%Only GPT-5.5 measured
Excel Modeling Benchmark (Vals AI)Not measured64.5%Only GPT-5.5 measured
Time Horizon Index (Vals AI)Not measured8.3%Only GPT-5.5 measured
τ²-Bench Telecom87.1%93.9%GPT-5.5
APEX-Agents52.4%55.1%GPT-5.5
METR time horizons74.3%Not measuredOnly GPT-5.4 measured
Remote Labor IndexNot measured6.3%Only GPT-5.5 measured
OSWorld 2.0Not measured13.0%Only GPT-5.5 measured
DeepResearch Bench35.1%54.0%GPT-5.5
PostTrainBench19.0%27.2%GPT-5.5
Vending-Bench 2$6,144$7,524GPT-5.5
Neither model has a result on the other 15 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index313
Reasoning benchmarks for GPT-5.4 and GPT-5.5
BenchmarkGPT-5.4GPT-5.5Leader
GPQA Diamond91.7%93.2%GPT-5.5
LiveBench reasoningNot measured89.7%Only GPT-5.5 measured
MMMU Pro (Vals AI)87.5%88.3%GPT-5.5
GPQA Diamond (Artificial Analysis run)92.0%93.5%GPT-5.5
Humanity's Last Exam (Artificial Analysis run)43.7%45.8%GPT-5.5
GPQA Diamond (Epoch AI run)93.3%90.7%GPT-5.4
ARC-AGI93.7%95.0%GPT-5.5
ARC-AGI-274.0%85.0%GPT-5.5
SimpleBenchNot measured69.0%Only GPT-5.5 measured
Chess puzzles48.0%63.0%GPT-5.5
Mystery game puzzles37.0%56.0%GPT-5.5
EBR-bench25.4%34.3%GPT-5.5
LMArena Text: hard prompts14851489GPT-5.5
LMArena Text: expert15071507GPT-5.5
LMArena Text: hard prompts in English14791483GPT-5.5
LMArena Vision: diagrams13271325GPT-5.4
LMArena Vision: homework13361332GPT-5.4
Kagi LLM Benchmark63.8%88.8%GPT-5.5
Neither model has a result on the other 4 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index42
Math benchmarks for GPT-5.4 and GPT-5.5
BenchmarkGPT-5.4GPT-5.5Leader
AIME96.7%Not measuredOnly GPT-5.4 measured
OTIS Mock AIME97.8%84.4%GPT-5.4
FrontierMath, tiers 1 to 378.6%85.3%GPT-5.5
FrontierMath, tier 449.0%72.5%GPT-5.5
ProofBench56.0%50.0%GPT-5.4
LMArena Text: math14881487GPT-5.4
LMArena Text: mathematical work14971496GPT-5.4
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index012
Knowledge benchmarks for GPT-5.4 and GPT-5.5
BenchmarkGPT-5.4GPT-5.5Leader
MMLU Pro87.5%88.1%GPT-5.5
LegalBench (Vals AI)86.0%86.5%GPT-5.5
CaseLaw (Vals AI)63.8%66.2%GPT-5.5
CorpFin (Vals AI)65.3%68.4%GPT-5.5
TaxEval (Vals AI)74.0%75.0%GPT-5.5
MortgageTax (Vals AI)68.3%68.8%GPT-5.5
MedQA (Vals AI)96.1%Not measuredOnly GPT-5.4 measured
MedCode (Vals AI)41.3%49.1%GPT-5.5
SimpleQA Verified45.1%63.0%GPT-5.5
LMArena Text: science14791485GPT-5.5
LMArena Text: medicine and healthcare14581465GPT-5.5
LMArena Text: legal and government14861491GPT-5.5
LMArena Text: business and finance14741478GPT-5.5
FACTS Benchmark SuiteNot measured66.7%Only GPT-5.5 measured
Neither model has a result on the other 4 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index08
Web design benchmarks for GPT-5.4 and GPT-5.5
BenchmarkGPT-5.4GPT-5.5Leader
Design Arena: websites11991256GPT-5.5
Design Arena: UI components12151260GPT-5.5
Design Arena: game dev11951294GPT-5.5
Design Arena: data viz11961265GPT-5.5
Design Arena: 3D10661198GPT-5.5
Design Arena: all code categories11801260GPT-5.5
Design Arena: SVG11421234GPT-5.5
Design Arena: ASCII art11881255GPT-5.5
LiveBench designNot measured70.3%Only GPT-5.5 measured
GeneralLike-for-likePublic boards, shown as published, kept out of the index825
General benchmarks for GPT-5.4 and GPT-5.5
BenchmarkGPT-5.4GPT-5.5Leader
LiveBench data analysisNot measured81.6%Only GPT-5.5 measured
Public Benefits Bench (Vals AI)Not measured60.9%Only GPT-5.5 measured
SAGE (Vals AI)43.3%51.5%GPT-5.5
MedScribe (Vals AI)77.5%86.9%GPT-5.5
Vals Multimodal IndexNot measured68.1%Only GPT-5.5 measured
IFBench73.9%75.9%GPT-5.5
Long context reasoning (AA-LCR)82.0%84.3%GPT-5.5
DTBench94.4%96.0%GPT-5.5
LMCA52.0%54.3%GPT-5.5
CL-bench27.9%Not measuredOnly GPT-5.4 measured
Surface Evolver BenchNot measured88.1%Only GPT-5.5 measured
Furniture assembly37.5%44.2%GPT-5.5
ExploitBenchNot measured47.4%Only GPT-5.5 measured
CL-bench Life21.7%22.2%GPT-5.5
LMArena Text14691471GPT-5.5
LMArena Text: instruction following14691478GPT-5.5
LMArena Text: creative writing14391455GPT-5.5
LMArena Vision13031297GPT-5.4
LMArena Text: English14671470GPT-5.5
LMArena Text: non-English14651467GPT-5.5
LMArena Text: Chinese15201534GPT-5.5
LMArena Text: German14721480GPT-5.5
LMArena Text: Japanese14851499GPT-5.5
LMArena Text: Korean14491460GPT-5.5
LMArena Text: French14931485GPT-5.4
LMArena Text: Spanish14521468GPT-5.5
LMArena Text: Russian14801473GPT-5.4
LMArena Text: Polish14721465GPT-5.4
LMArena Text: multi-turn14821476GPT-5.4
LMArena Text: longer queries14731484GPT-5.5
LMArena Text: writing and language14581466GPT-5.5
LMArena Text: entertainment and media14381449GPT-5.5
LMArena Vision: English13011292GPT-5.4
LMArena Vision: Chinese13541360GPT-5.5
LMArena Vision: reading text13121306GPT-5.4
LMArena Vision: humor12841293GPT-5.5
LMArena Vision: recognizing things12511231GPT-5.4
LMArena Vision: creative writing12831301GPT-5.5
LMArena Document14711486GPT-5.5
Neither model has a result on the other 7 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5.4 and GPT-5.5
BenchmarkGPT-5.4GPT-5.5Leader
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

GPT-5.4$0.010Fits in one request
GPT-5.5$0.020Fits in one request

GPT-5.4 costs less

Repository review

50K input + 3K output tokens

GPT-5.4$0.170Fits in one request
GPT-5.5$0.340Fits in one request

GPT-5.4 costs less

Agent loop

200K input + 10K output tokens

GPT-5.4$0.650Fits in one request
GPT-5.5$1.30Fits in one request

GPT-5.4 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 GPT-5.4 and GPT-5.5
SpecificationOpenAI GPT-5.4OpenAI GPT-5.5
MakerOpenAIOpenAI
WeightsNo public weights listedNo public weights listed
ReleasedMarch 5, 2026April 24, 2026
Context window1.05M1.05M
Input $/Mtok$2.50$5.00
Output $/Mtok$15.00$30.00
Coding index64.471.3
Independent measurements behind it35
Cost per solved task$1.02$1.65
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5.4 better than GPT-5.5 for coding?

Pick GPT-5.5 for coding: it scores 71.3 on the Vibe Coding Index against 64.4 for GPT-5.4. Pick GPT-5.4 for output tokens 2.0x cheaper ($15.00 vs $30.00 per million) and a lower cost per solved task ($1.02 against $1.65).

Which is cheaper, GPT-5.4 or GPT-5.5?

GPT-5.4 costs $2.50 per million input tokens and $15.00 per million output tokens; GPT-5.5 costs $5.00 and $30.00. Per solved SWE-bench Verified task, GPT-5.4 costs $1.02 and GPT-5.5 $1.65.

Which has the bigger context window, GPT-5.4 or GPT-5.5?

Neither: both take 1.05M tokens.

How many benchmarks is the GPT-5.4 vs GPT-5.5 comparison based on?

106 benchmarks have a result for both models as of October 21, 2026. GPT-5.4's coding index rests on 3 independent benchmarks and GPT-5.5'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/gpt-5-4-vs-gpt-5-5" title="GPT-5.4 vs GPT-5.5, 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/gpt-5-4-vs-gpt-5-5">LLMBenchmark.ai, GPT-5.4 vs GPT-5.5</a></p>