Noometry

Model comparison

GPT-4o vs GPT-5.5

GPT-5.5 is the stronger model overall, scoring 63.4 to 28.6 on the Noometry Index. GPT-4o costs 2.6× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

Last verified . 37 shared benchmarks.

GPT-4o OpenAI

28.6

Rank #324 Confirmed

GPT-5.5 OpenAI

63.4

Rank #9 Confirmed

Summary

  • They share 37 benchmarks with published results for both. GPT-4o scores higher in 0 categories and GPT-5.5 in 10 categories; 10 gaps are clear of the uncertainty.
  • The widest gap is in math, where GPT-5.5 leads 81.7 to 10.6.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 6.4% for GPT-4o and 100% for GPT-5.5.
  • GPT-4o is cheaper at $2.50 / $10 per million input/output tokens, against $5 / $30 for GPT-5.5.
  • GPT-5.5 accepts more context: 1.05M tokens versus 128K.

Side by side

GPT-4oGPT-5.5
ProviderOpenAIOpenAI
Noometry Index28.663.4
Released2024-05-132026-04-23
WeightsProprietaryProprietary
Context window128K1.05M
Max output16K128K
Input $ / M tokens$2.50$5
Output $ / M tokens$10$30
Results tracked7271

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding GPT-5.5 leads

GPT-4o: 24.8 (#328), GPT-5.5: 58.2 (#17)

BenchmarkGPT-4oGPT-5.5
SWE-bench Verified31%80.6%
GSO0%40.2%
WeirdML25.1%84.9%
LMArena Coding12971494
DeepSWE—67%
FrontierCode—43%
SWE-bench Verified (bash only)21.6%—
Aider Polyglot45.3%—
LMArena WebDev—1513
SciCode—56.1%
BigCodeBench Instruct51.1%—
LiveBench Coding51.4%—
MirrorCode—10%
BigCodeBench Complete61.1%—
CadEval26%—
ALE-Bench—1,943
HumanEval+87.2%—
MBPP+72.2%—

Agentic & Tool Use GPT-5.5 leads

GPT-4o: 21.0 (#141), GPT-5.5: 50.7 (#6)

BenchmarkGPT-4oGPT-5.5
LMArena Search10061242
Terminal-Bench—84.7%
APEX-Agents—55.1%
OSWorld 2.0—13%
GDPval9.9%—
Remote Labor Index—6.3%
TheAgentCompany8.6%—
τ²-bench Banking—44.6%
Cybench12.5%—
DeepResearch Bench—54%
PostTrainBench—27.2%
BALROG32.3%—
ExploitBench—47.4%
GBAEval—53.2%
GDP.pdf—26%
METR Time Horizons40.8%—
Vending-Bench 2—7,524

Reasoning GPT-5.5 leads

GPT-4o: 9.4 (#343), GPT-5.5: 72.8 (#11)

BenchmarkGPT-4oGPT-5.5
ARC-AGI-20%85%
SimpleBench17.8%69%
ARC-AGI-14.5%95%
CritPt0%27.1%
Chess Puzzles13%54%
LMArena Hard Prompts12811489
DTBench64.5%96%
LMCA16.6%54.3%
Epoch Capabilities Index128.97159.1
ForecastBench57.760.6
Kagi LLM Benchmark—88.8%
NYT Connections (extended)—96.2%
EnigmaEval0.8%—
EBR-Bench—34.3%
LiveBench Reasoning55.8%—
Mystery Game Puzzles—56%
LiveBench Data Analysis60.9%—
Surface Evolver Bench—88.1%
Bench to the Future 3—0.14
LiveBench55.3%—

Math GPT-5.5 leads

GPT-4o: 10.6 (#312), GPT-5.5: 81.7 (#11)

BenchmarkGPT-4oGPT-5.5
FrontierMath (Tiers 1-3)0.4%85.3%
OTIS Mock AIME 2024-20256.4%100%
LMArena Math12851486
FrontierMath (Feb 2025 set)0.3%51.7%
FrontierMath Tier 4—72.5%
MathArena Final-Answer Competitions—94.3%
ProofBench—50%
Omni-MATH29.3%—
LiveBench Math49.5%—
MATH Level 553.3%—
FrontierMath Erdős—0%
FrontierMath Tier 4 (v1)—35.4%

Knowledge GPT-5.5 leads

GPT-4o: 28.8 (#242), GPT-5.5: 64.4 (#17)

BenchmarkGPT-4oGPT-5.5
GPQA Diamond49.2%94%
SimpleQA Verified26%63%
Vectara Hallucination Rate9.6%9.3%
LMArena Expert12501508
Humanity's Last Exam2.7%—
MMLU-Pro71.3%—
Confabulations15.3%—
GPQA (HELM)52%—
MMLU88.1%—

Multimodal GPT-5.5 leads

GPT-4o: 34.5 (#91), GPT-5.5: 46.9 (#12)

BenchmarkGPT-4oGPT-5.5
LMArena Vision11371297
Video-MME71.9%—
GeoBench71%—
VPCT40%—
Blueprint-Bench 2—36.2%
Furniture Assembly—44.2%
LMArena Document—1486
ScienceQA88.5%—

Multilingual GPT-5.5 leads

GPT-4o: 43.2 (#186), GPT-5.5: 56.4 (#20)

BenchmarkGPT-4oGPT-5.5
LMArena Non-English12831467
LMArena Chinese12771533
LMArena French13041486
LMArena German12821480
LMArena Japanese12571498
LMArena Korean12341460
LMArena Russian12861473
LMArena Spanish12921468

Instruction Following GPT-5.5 leads

GPT-4o: 66.6 (#207), GPT-5.5: 77.5 (#18)

BenchmarkGPT-4oGPT-5.5
LMArena Instruction Following12781479
LiveBench Instruction Following68.6%—
IFEval81.7%—

Long Context GPT-5.5 leads

GPT-4o: 39.4 (#179), GPT-5.5: 48.3 (#12)

BenchmarkGPT-4oGPT-5.5
LMArena Longer Query12891484
Fiction.LiveBench66.7%—
CL-bench Life—22.2%

Writing & Preference GPT-5.5 leads

GPT-4o: 52.6 (#166), GPT-5.5: 72.7 (#13)

BenchmarkGPT-4oGPT-5.5
LMArena Text13001472
LMArena Creative Writing12921455
LMArena Multi-Turn13021476
Short-Story Creative Writing81.8%—
EQ-Bench Creative Writing—1844
WildBench82.8%—
EQ-Bench 4—1315
LiveBench Language47.6%—

Frequently asked questions

Is GPT-4o better than GPT-5.5?

GPT-5.5 is the stronger model overall, scoring 63.4 to 28.6 on the Noometry Index. GPT-4o costs 2.6× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

Which is cheaper, GPT-4o or GPT-5.5?

GPT-4o is cheaper. It lists at $2.50 per million input tokens and $10 per million output tokens; GPT-5.5 lists at $5 and $30.

Is GPT-4o or GPT-5.5 better for coding?

GPT-5.5 scores higher on coding benchmarks: 58.2 versus 24.8 in the Noometry coding category.

Which has the bigger context window?

GPT-5.5 does, with 1.05M tokens against 128K.

How many benchmarks do GPT-4o and GPT-5.5 share?

37 benchmarks have published results for both models. GPT-4o has 72 scored results on Noometry and GPT-5.5 has 71.

Related comparisons

Go deeper