Noometry

Model comparison

Claude Opus 4.6 vs GPT-6 Astra

GPT-6 Astra is the stronger model overall, scoring 70.8 to 58.2 on the Noometry Index. Claude Opus 4.6 costs 2.0× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.

Last verified . 46 shared benchmarks.

Claude Opus 4.6 Anthropic

58.2

Rank #20 Confirmed

GPT-6 Astra OpenAI

70.8

Rank #1 Confirmed

Summary

  • They share 46 benchmarks with published results for both. Claude Opus 4.6 scores higher in 3 categories and GPT-6 Astra in 7 categories; 10 gaps are clear of the uncertainty.
  • The widest gap is in math, where GPT-6 Astra leads 93.5 to 63.0.
  • The biggest single-benchmark swing is FrontierMath Tier 4: 26.8% for Claude Opus 4.6 and 97.6% for GPT-6 Astra.
  • Claude Opus 4.6 is cheaper at $5 / $25 per million input/output tokens, against $10 / $50 for GPT-6 Astra.
  • GPT-6 Astra accepts more context: 1.05M tokens versus 1M.

Side by side

Claude Opus 4.6GPT-6 Astra
ProviderAnthropicOpenAI
Noometry Index58.270.8
Released2026-02-042026-09-03
WeightsProprietaryProprietary
Context window1M1.05M
Max output128K128K
Input $ / M tokens$5$10
Output $ / M tokens$25$50
Results tracked6856

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

Category by category

Coding GPT-6 Astra leads

Claude Opus 4.6: 57.2 (#20), GPT-6 Astra: 73.7 (#2)

BenchmarkClaude Opus 4.6GPT-6 Astra
FrontierCode26.6%53.3%
LMArena WebDev15471786
GSO41.2%79.4%
WeirdML78%93.6%
LMArena Coding15361487
ALE-Bench996.52,951
SWE-bench Verified78.7%—
DeepSWE—74.1%
SWE-bench Verified (bash only)75.6%—
SWE-bench Multilingual72%—
FrontierSWE—65.5%
SciCode—56.5%
MirrorCode—46.7%
AlgoTune1.47—

Agentic & Tool Use GPT-6 Astra leads

Claude Opus 4.6: 51.1 (#4), GPT-6 Astra: 52.9 (#3)

BenchmarkClaude Opus 4.6GPT-6 Astra
APEX-Agents46.3%64.7%
Remote Labor Index4.2%20.8%
Vending-Bench 28,01815,515
Terminal-Bench79.8%—
τ²-bench Banking27.3%—
Cybench93%—
DeepResearch Bench55.3%—
BALROG—68.3%
GBAEval44.1%—
GDP.pdf—34.2%
LMArena Search1253—
METR Time Horizons78.9%—

Reasoning GPT-6 Astra leads

Claude Opus 4.6: 57.8 (#23), GPT-6 Astra: 85.1 (#1)

BenchmarkClaude Opus 4.6GPT-6 Astra
ARC-AGI-269.2%95%
NYT Connections (extended)92.1%98.1%
ARC-AGI-194%98.5%
Chess Puzzles17%72%
EBR-Bench12.7%76.2%
LMArena Hard Prompts15271462
Mystery Game Puzzles25%84%
DTBench91.2%97.3%
LMCA55.8%64.4%
Epoch Capabilities Index155.24166.45
SimpleBench67.6%—
Kagi LLM Benchmark83.6%—
CritPt—31.7%
EnigmaEval7.6%—
Thematic Generalization80.6%—
Bench to the Future 3—0.14
ForecastBench60—

Math GPT-6 Astra leads

Claude Opus 4.6: 63.0 (#31), GPT-6 Astra: 93.5 (#2)

BenchmarkClaude Opus 4.6GPT-6 Astra
FrontierMath (Tiers 1-3)66%93.7%
FrontierMath Tier 426.8%97.6%
OTIS Mock AIME 2024-202594.4%100%
ProofBench50%99%
LMArena Math15191465
MathArena Final-Answer Competitions78.5%—
FrontierMath (Feb 2025 set)40.7%—
FrontierMath Erdős—2.9%
FrontierMath Tier 4 (v1)22.9%—

Knowledge GPT-6 Astra leads

Claude Opus 4.6: 61.9 (#26), GPT-6 Astra: 75.3 (#1)

BenchmarkClaude Opus 4.6GPT-6 Astra
GPQA Diamond90.5%95.8%
Humanity's Last Exam34.4%54.8%
SimpleQA Verified47%75.6%
Vectara Hallucination Rate12.2%8.7%
LMArena Expert15461483

Multimodal GPT-6 Astra leads

Claude Opus 4.6: 37.3 (#74), GPT-6 Astra: 55.0 (#3)

BenchmarkClaude Opus 4.6GPT-6 Astra
LMArena Vision13161281
Furniture Assembly28.3%80%
LMArena Document15071468
Blueprint-Bench 2—49.7%

Multilingual Claude Opus 4.6 leads

Claude Opus 4.6: 57.9 (#6), GPT-6 Astra: 53.7 (#61)

BenchmarkClaude Opus 4.6GPT-6 Astra
LMArena Non-English14891430
LMArena Chinese15511484
LMArena French15131456
LMArena German15021440
LMArena Japanese14841379
LMArena Korean14641426
LMArena Russian14971436
LMArena Spanish15101407

Instruction Following Claude Opus 4.6 leads

Claude Opus 4.6: 79.5 (#4), GPT-6 Astra: 76.3 (#44)

BenchmarkClaude Opus 4.6GPT-6 Astra
LMArena Instruction Following15231450

Long Context Claude Opus 4.6 leads

Claude Opus 4.6: 48.1 (#13), GPT-6 Astra: 44.5 (#62)

BenchmarkClaude Opus 4.6GPT-6 Astra
LMArena Longer Query15201456
CL-bench20.7%—
CL-bench Life17%—

Writing & Preference GPT-6 Astra leads

Claude Opus 4.6: 73.5 (#10), GPT-6 Astra: 75.3 (#7)

BenchmarkClaude Opus 4.6GPT-6 Astra
LMArena Text15031441
LMArena Creative Writing15051418
EQ-Bench Creative Writing18092173
LMArena Multi-Turn15131448
EQ-Bench 41223—

Frequently asked questions

Is Claude Opus 4.6 better than GPT-6 Astra?

GPT-6 Astra is the stronger model overall, scoring 70.8 to 58.2 on the Noometry Index. Claude Opus 4.6 costs 2.0× less per token, which makes it the better buy when GPT-6 Astra's lead doesn't matter for your workload.

Which is cheaper, Claude Opus 4.6 or GPT-6 Astra?

Claude Opus 4.6 is cheaper. It lists at $5 per million input tokens and $25 per million output tokens; GPT-6 Astra lists at $10 and $50.

Is Claude Opus 4.6 or GPT-6 Astra better for coding?

GPT-6 Astra scores higher on coding benchmarks: 73.7 versus 57.2 in the Noometry coding category.

Which has the bigger context window?

GPT-6 Astra does, with 1.05M tokens against 1M.

How many benchmarks do Claude Opus 4.6 and GPT-6 Astra share?

46 benchmarks have published results for both models. Claude Opus 4.6 has 68 scored results on Noometry and GPT-6 Astra has 56.

Related comparisons

Go deeper