Noometry

Model comparison

Claude Opus 4.8 vs GPT-5.2

Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 54.1 on the Noometry Index. GPT-5.2 costs 2.1× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.

Last verified . 52 shared benchmarks.

Claude Opus 4.8 Anthropic

60.7

Rank #9 Confirmed

GPT-5.2 OpenAI

54.1

Rank #31 Confirmed

Summary

  • They share 52 benchmarks with published results for both. Claude Opus 4.8 scores higher in 9 categories and GPT-5.2 in 1 category; 10 gaps are clear of the uncertainty.
  • The widest gap is in math, where Claude Opus 4.8 leads 78.2 to 59.8.
  • The biggest single-benchmark swing is ProofBench: 69% for Claude Opus 4.8 and 15% for GPT-5.2.
  • GPT-5.2 is cheaper at $1.75 / $14 per million input/output tokens, against $5 / $25 for Claude Opus 4.8.
  • Claude Opus 4.8 accepts more context: 1M tokens versus 400K.

Side by side

Claude Opus 4.8GPT-5.2
ProviderAnthropicOpenAI
Noometry Index60.754.1
Released2026-05-282025-12-11
WeightsProprietaryProprietary
Context window1M400K
Max output128K128K
Input $ / M tokens$5$1.75
Output $ / M tokens$25$14
Results tracked6567

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

Category by category

Coding Claude Opus 4.8 leads

Claude Opus 4.8: 60.2 (#9), GPT-5.2: 52.2 (#32)

BenchmarkClaude Opus 4.8GPT-5.2
LMArena WebDev15561415
GSO47.1%27.4%
WeirdML82.9%72.2%
LMArena Coding14911447
ALE-Bench1,5641,294
SWE-bench Verified—73.8%
DeepSWE59%—
FrontierCode46.5%—
SWE-bench Verified (bash only)—72.8%
SWE-bench Multilingual—66.7%
SciCode53.5%—
AlgoTune—2.05

Agentic & Tool Use Claude Opus 4.8 leads

Claude Opus 4.8: 47.6 (#10), GPT-5.2: 40.0 (#24)

BenchmarkClaude Opus 4.8GPT-5.2
Remote Labor Index8.3%2.5%
τ²-bench Banking39.7%32.2%
DeepResearch Bench50.2%41.1%
LMArena Search12041207
Vending-Bench 25,7873,591
Terminal-Bench—64.9%
APEX-Agents48.9%—
Berkeley Function Calling Leaderboard—55.9%
OSWorld 2.020.6%—
GDPval—49.7%
τ²-bench Airline—83%
τ²-bench Retail—81.6%
τ²-bench Telecom—89.7%
PostTrainBench33.8%—
GBAEval70.9%—
GDP.pdf24%—
METR Time Horizons—75.3%

Reasoning Claude Opus 4.8 leads

Claude Opus 4.8: 64.6 (#14), GPT-5.2: 50.1 (#33)

BenchmarkClaude Opus 4.8GPT-5.2
ARC-AGI-272.1%52.9%
SimpleBench64.8%45.8%
Kagi LLM Benchmark88.8%73.3%
NYT Connections (extended)91.1%83.6%
ARC-AGI-192.5%86.2%
Chess Puzzles34%49%
EnigmaEval23.5%10.4%
EBR-Bench28.6%23%
LMArena Hard Prompts14811445
Mystery Game Puzzles36%23%
DTBench94.9%90.9%
LMCA57.5%43.9%
Epoch Capabilities Index158.3153.45
ForecastBench59.960.1
CritPt20.9%—
Surface Evolver Bench87.5%—
Bench to the Future 30.14—

Math Claude Opus 4.8 leads

Claude Opus 4.8: 78.2 (#11), GPT-5.2: 59.8 (#32)

BenchmarkClaude Opus 4.8GPT-5.2
FrontierMath (Tiers 1-3)80%67.4%
FrontierMath Tier 456.1%31.7%
MathArena Final-Answer Competitions91.8%72%
OTIS Mock AIME 2024-202598.3%96.1%
ProofBench69%15%
LMArena Math14871439
FrontierMath (Feb 2025 set)47.2%40.7%
FrontierMath Tier 4 (v1)31.3%18.8%

Knowledge Claude Opus 4.8 leads

Claude Opus 4.8: 61.3 (#26), GPT-5.2: 59.2 (#29)

BenchmarkClaude Opus 4.8GPT-5.2
GPQA Diamond91%91.4%
SimpleQA Verified53%37.1%
LMArena Expert15011445
Humanity's Last Exam—27.8%
Vectara Hallucination Rate—8.4%

Multimodal GPT-5.2 leads

Claude Opus 4.8: 43.2 (#20), GPT-5.2: 50.9 (#7)

BenchmarkClaude Opus 4.8GPT-5.2
LMArena Vision12951268
Furniture Assembly42.5%38.3%
LMArena Document14751405
VPCT—84%
Blueprint-Bench 214.5%—

Multilingual Claude Opus 4.8 leads

Claude Opus 4.8: 55.4 (#31), GPT-5.2: 53.6 (#63)

BenchmarkClaude Opus 4.8GPT-5.2
LMArena Non-English14501425
LMArena Chinese15091460
LMArena French14821456
LMArena German14701448
LMArena Japanese14401421
LMArena Korean14321390
LMArena Russian14761440
LMArena Spanish14701433

Instruction Following Claude Opus 4.8 leads

Claude Opus 4.8: 77.6 (#22), GPT-5.2: 74.9 (#83)

BenchmarkClaude Opus 4.8GPT-5.2
LMArena Instruction Following14761417

Long Context Claude Opus 4.8 leads

Claude Opus 4.8: 45.4 (#37), GPT-5.2: 44.0 (#74)

BenchmarkClaude Opus 4.8GPT-5.2
LMArena Longer Query14821428
CL-bench—18.2%

Writing & Preference Claude Opus 4.8 leads

Claude Opus 4.8: 72.1 (#17), GPT-5.2: 67.0 (#30)

BenchmarkClaude Opus 4.8GPT-5.2
LMArena Text14601439
LMArena Creative Writing14531402
EQ-Bench Creative Writing18401703
LMArena Multi-Turn14761458
EQ-Bench 41281—

Frequently asked questions

Is Claude Opus 4.8 better than GPT-5.2?

Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 54.1 on the Noometry Index. GPT-5.2 costs 2.1× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.

Which is cheaper, Claude Opus 4.8 or GPT-5.2?

GPT-5.2 is cheaper. It lists at $1.75 per million input tokens and $14 per million output tokens; Claude Opus 4.8 lists at $5 and $25.

Is Claude Opus 4.8 or GPT-5.2 better for coding?

Claude Opus 4.8 scores higher on coding benchmarks: 60.2 versus 52.2 in the Noometry coding category.

Which has the bigger context window?

Claude Opus 4.8 does, with 1M tokens against 400K.

How many benchmarks do Claude Opus 4.8 and GPT-5.2 share?

52 benchmarks have published results for both models. Claude Opus 4.8 has 65 scored results on Noometry and GPT-5.2 has 67.

Related comparisons

Go deeper