Noometry

Model comparison

Claude Opus 5 vs GPT-5.5 Instant

Claude Opus 5 is the stronger model overall, scoring 67.7 to 42.5 on the Noometry Index.

Last verified . 27 shared benchmarks.

Claude Opus 5 Anthropic

67.7

Rank #4 Confirmed

GPT-5.5 Instant OpenAI

42.5

Rank #99 Confirmed

Summary

  • They share 27 benchmarks with published results for both. Claude Opus 5 scores higher in 9 categories and GPT-5.5 Instant in 0 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in math, where Claude Opus 5 leads 86.1 to 26.1.
  • The biggest single-benchmark swing is FrontierMath Tier 4: 73.2% for Claude Opus 5 and 2.4% for GPT-5.5 Instant.

Side by side

Claude Opus 5GPT-5.5 Instant
ProviderAnthropicOpenAI
Noometry Index67.742.5
Released2026-07-242026-05-05
WeightsProprietaryProprietary
Context window1M—
Max output128K—
Input $ / M tokens$5—
Output $ / M tokens$25—
Results tracked5727

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

Category by category

Coding Claude Opus 5 leads

Claude Opus 5: 67.0 (#5), GPT-5.5 Instant: 43.7 (#73)

BenchmarkClaude Opus 5GPT-5.5 Instant
SciCode56.4%48.6%
LMArena Coding15331433
DeepSWE73.6%—
FrontierCode53.4%—
CursorBench46.6%—
LMArena WebDev1694—
FrontierSWE52%—
WeirdML91.8%—
ALE-Bench2,165—

Agentic & Tool Use Not comparable

Claude Opus 5: 55.5 (#1), GPT-5.5 Instant: —

BenchmarkClaude Opus 5GPT-5.5 Instant
APEX-Agents65.8%—
OSWorld 2.031.4%—
τ²-bench Banking48.7%—
PostTrainBench35%—
BALROG63.4%—
GBAEval79.6%—
GDP.pdf24%—
Vending-Bench 211,182—

Reasoning Claude Opus 5 leads

Claude Opus 5: 77.1 (#3), GPT-5.5 Instant: 24.2 (#149)

BenchmarkClaude Opus 5GPT-5.5 Instant
CritPt29.1%0%
Chess Puzzles42%12%
LMArena Hard Prompts15251426
Epoch Capabilities Index162.67142.48
ARC-AGI-290.4%—
SimpleBench80.6%—
NYT Connections (extended)94.3%—
ARC-AGI-197.5%—
EBR-Bench45.7%—
Mystery Game Puzzles59%—
DTBench97.9%—
LMCA64.5%—
Bench to the Future 30.12—

Math Claude Opus 5 leads

Claude Opus 5: 86.1 (#6), GPT-5.5 Instant: 26.1 (#241)

BenchmarkClaude Opus 5GPT-5.5 Instant
FrontierMath (Tiers 1-3)85.6%26.3%
FrontierMath Tier 473.2%2.4%
OTIS Mock AIME 2024-202598.9%68.1%
LMArena Math15331420
ProofBench99%—

Knowledge Claude Opus 5 leads

Claude Opus 5: 66.8 (#8), GPT-5.5 Instant: 48.9 (#68)

BenchmarkClaude Opus 5GPT-5.5 Instant
GPQA Diamond93.9%82.5%
LMArena Expert15581409
SimpleQA Verified59.9%—

Multimodal Claude Opus 5 leads

Claude Opus 5: 51.1 (#6), GPT-5.5 Instant: 39.3 (#48)

BenchmarkClaude Opus 5GPT-5.5 Instant
LMArena Vision13211250
LMArena Document15161403
Blueprint-Bench 230.4%—
Furniture Assembly60.8%—

Multilingual Claude Opus 5 leads

Claude Opus 5: 58.8 (#3), GPT-5.5 Instant: 53.0 (#76)

BenchmarkClaude Opus 5GPT-5.5 Instant
LMArena Non-English15001417
LMArena Chinese15751457
LMArena French15221428
LMArena German15221410
LMArena Japanese15101408
LMArena Korean15261393
LMArena Russian15071430
LMArena Spanish15191430

Instruction Following Claude Opus 5 leads

Claude Opus 5: 79.3 (#5), GPT-5.5 Instant: 74.4 (#94)

BenchmarkClaude Opus 5GPT-5.5 Instant
LMArena Instruction Following15161407

Long Context Claude Opus 5 leads

Claude Opus 5: 46.5 (#20), GPT-5.5 Instant: 43.4 (#92)

BenchmarkClaude Opus 5GPT-5.5 Instant
LMArena Longer Query15161422

Writing & Preference Claude Opus 5 leads

Claude Opus 5: 79.3 (#2), GPT-5.5 Instant: 62.0 (#81)

BenchmarkClaude Opus 5GPT-5.5 Instant
LMArena Text15061419
LMArena Creative Writing14911420
LMArena Multi-Turn14961434
EQ-Bench Creative Writing2133—
EQ-Bench 41385—

Frequently asked questions

Is Claude Opus 5 better than GPT-5.5 Instant?

Claude Opus 5 is the stronger model overall, scoring 67.7 to 42.5 on the Noometry Index.

Is Claude Opus 5 or GPT-5.5 Instant better for coding?

Claude Opus 5 scores higher on coding benchmarks: 67.0 versus 43.7 in the Noometry coding category.

How many benchmarks do Claude Opus 5 and GPT-5.5 Instant share?

27 benchmarks have published results for both models. Claude Opus 5 has 57 scored results on Noometry and GPT-5.5 Instant has 27.

Related comparisons

Go deeper