Noometry

Model comparison

DeepSeek-V3 vs GPT-5.5

GPT-5.5 is the stronger model overall, scoring 63.4 to 39.5 on the Noometry Index. DeepSeek-V3 costs 28× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

Last verified . 31 shared benchmarks.

DeepSeek-V3 DeepSeek

39.5

Rank #166 Confirmed

GPT-5.5 OpenAI

63.4

Rank #9 Confirmed

Summary

  • They share 31 benchmarks with published results for both. DeepSeek-V3 scores higher in 0 categories and GPT-5.5 in 8 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where GPT-5.5 leads 72.8 to 20.5.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 37.8% for DeepSeek-V3 and 100% for GPT-5.5.
  • DeepSeek-V3 is cheaper at $0.24 / $0.90 per million input/output tokens, against $5 / $30 for GPT-5.5.
  • GPT-5.5 accepts more context: 1.05M tokens versus 164K.
  • DeepSeek-V3 has downloadable open weights; the other is API-only.

Side by side

DeepSeek-V3GPT-5.5
ProviderDeepSeekOpenAI
Noometry Index39.563.4
Released2024-12-262026-04-23
WeightsOpenProprietary
Context window164K1.05M
Max output164K128K
Input $ / M tokens$0.24$5
Output $ / M tokens$0.90$30
Results tracked6071

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

Category by category

Coding GPT-5.5 leads

DeepSeek-V3: 42.3 (#106), GPT-5.5: 58.2 (#17)

BenchmarkDeepSeek-V3GPT-5.5
SciCode35.8%56.1%
WeirdML36.1%84.9%
LMArena Coding13681494
SWE-bench Verified—80.6%
DeepSWE—67%
FrontierCode—43%
Aider Polyglot55.1%—
LMArena WebDev—1513
GSO—40.2%
BigCodeBench Instruct50%—
LiveBench Coding70.9%—
MirrorCode—10%
BigCodeBench Complete62.2%—
ALE-Bench—1,943
HumanEval+86.6%—
MBPP+73%—

Agentic & Tool Use Not comparable

DeepSeek-V3: —, GPT-5.5: 50.7 (#6)

BenchmarkDeepSeek-V3GPT-5.5
Terminal-Bench—84.7%
APEX-Agents—55.1%
OSWorld 2.0—13%
Remote Labor Index—6.3%
τ²-bench Banking—44.6%
DeepResearch Bench—54%
PostTrainBench—27.2%
ExploitBench—47.4%
GBAEval—53.2%
GDP.pdf—26%
LMArena Search—1242
METR Time Horizons49.6%—
Vending-Bench 2—7,524

Reasoning GPT-5.5 leads

DeepSeek-V3: 20.5 (#236), GPT-5.5: 72.8 (#11)

BenchmarkDeepSeek-V3GPT-5.5
SimpleBench27.2%69%
Kagi LLM Benchmark52.3%88.8%
CritPt0%27.1%
LMArena Hard Prompts13651489
DTBench64.8%96%
LMCA15.5%54.3%
Epoch Capabilities Index135.94159.1
ForecastBench59.160.6
ARC-AGI-2—85%
NYT Connections (extended)—96.2%
ARC-AGI-1—95%
Chess Puzzles—54%
EBR-Bench—34.3%
LiveBench Reasoning65.8%—
Mystery Game Puzzles—56%
LiveBench Data Analysis60.9%—
Surface Evolver Bench—88.1%
Bench to the Future 3—0.14
BIG-Bench Hard87.5%—
HellaSwag88.9%—
LiveBench66.9%—
PIQA84.7%—
WinoGrande85.2%—

Math GPT-5.5 leads

DeepSeek-V3: 32.1 (#219), GPT-5.5: 81.7 (#11)

BenchmarkDeepSeek-V3GPT-5.5
OTIS Mock AIME 2024-202537.8%100%
LMArena Math13731486
FrontierMath (Feb 2025 set)1.7%51.7%
FrontierMath (Tiers 1-3)—85.3%
FrontierMath Tier 4—72.5%
MathArena Final-Answer Competitions—94.3%
ProofBench—50%
Omni-MATH40.3%—
LiveBench Math73.5%—
MATH Level 575.5%—
FrontierMath Erdős—0%
FrontierMath Tier 4 (v1)—35.4%

Knowledge GPT-5.5 leads

DeepSeek-V3: 37.5 (#155), GPT-5.5: 64.4 (#17)

BenchmarkDeepSeek-V3GPT-5.5
GPQA Diamond67.6%94%
Vectara Hallucination Rate6.1%9.3%
LMArena Expert13511508
SimpleQA Verified—63%
MMLU-Pro72.3%—
Confabulations26.1%—
GPQA (HELM)53.8%—
ARC (AI2) Challenge95.3%—
MMLU87.2%—
TriviaQA82.9%—

Multimodal Not comparable

DeepSeek-V3: —, GPT-5.5: 46.9 (#12)

BenchmarkDeepSeek-V3GPT-5.5
LMArena Vision—1297
Blueprint-Bench 2—36.2%
Furniture Assembly—44.2%
LMArena Document—1486

Multilingual GPT-5.5 leads

DeepSeek-V3: 48.5 (#143), GPT-5.5: 56.4 (#20)

BenchmarkDeepSeek-V3GPT-5.5
LMArena Non-English13581467
LMArena Chinese13911533
LMArena French13851486
LMArena German13741480
LMArena Japanese13331498
LMArena Korean13191460
LMArena Russian13731473
LMArena Spanish13581468

Instruction Following GPT-5.5 leads

DeepSeek-V3: 72.8 (#130), GPT-5.5: 77.5 (#18)

BenchmarkDeepSeek-V3GPT-5.5
LMArena Instruction Following13451479
LiveBench Instruction Following81.5%—
IFEval83.2%—

Long Context GPT-5.5 leads

DeepSeek-V3: 34.0 (#253), GPT-5.5: 48.3 (#12)

BenchmarkDeepSeek-V3GPT-5.5
LMArena Longer Query13521484
Fiction.LiveBench50%—
CL-bench Life—22.2%

Writing & Preference GPT-5.5 leads

DeepSeek-V3: 57.4 (#130), GPT-5.5: 72.7 (#13)

BenchmarkDeepSeek-V3GPT-5.5
LMArena Text13751472
LMArena Creative Writing13641455
EQ-Bench Creative Writing14721844
LMArena Multi-Turn13891476
Short-Story Creative Writing77%—
WildBench83%—
EQ-Bench 4—1315
LiveBench Language49.1%—

Frequently asked questions

Is DeepSeek-V3 better than GPT-5.5?

GPT-5.5 is the stronger model overall, scoring 63.4 to 39.5 on the Noometry Index. DeepSeek-V3 costs 28× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

Which is cheaper, DeepSeek-V3 or GPT-5.5?

DeepSeek-V3 is cheaper. It lists at $0.24 per million input tokens and $0.90 per million output tokens; GPT-5.5 lists at $5 and $30.

Is DeepSeek-V3 or GPT-5.5 better for coding?

GPT-5.5 scores higher on coding benchmarks: 58.2 versus 42.3 in the Noometry coding category.

Which has the bigger context window?

GPT-5.5 does, with 1.05M tokens against 164K.

How many benchmarks do DeepSeek-V3 and GPT-5.5 share?

31 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and GPT-5.5 has 71.

Related comparisons

Go deeper