SShogo Bench

Claude Opus 5 vs GPT-6 Astra

GPT-6 Astra leads Claude Opus 5 by 5.5 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Claude Opus 5 wins 5 and GPT-6 Astra wins 18 of 24 shared benchmarks. Claude Opus 5 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarClaude Opus 5GPT-6 Astra
Agentic coding7480
Long-horizon autonomy7480
Knowledge-work agents5256
Computer use5667
Security & context learning5963

Shared benchmarks

BenchmarkClaude Opus 5GPT-6 Astra
APEX-Agentsindex65.8%64.7%
DeepSWEindex73.6%74.1%
Vending-Bench 2index$11,182$15,515
FrontierSWEindex52.0%65.5%
PostTrainBenchindex35.7%41.9%
ARC-AGI97.5%98.5%
ARC-AGI-290.4%95.0%
BALROG63.4%68.3%
Chess Puzzles43.0%76.0%
CritPt29.1%31.7%
DTBench97.9%97.3%
FrontierCode53.4%53.3%
FrontierMath Tier 473.2%97.6%
FrontierMath Tiers 1-385.6%93.7%
GPQA Diamond93.9%95.8%
LMCA64.5%64.4%
MirrorCode87.8%60.0%
OTIS Mock AIME98.9%100.0%
ProofBench99.0%99.0%
SciCode56.4%56.5%
SimpleBench80.6%86.5%
SimpleQA Verified59.9%75.6%
WebDev Arena16871800
WeirdML v319.0%42.2%

Specs and price

Claude Opus 5GPT-6 Astra
Agent Index65.4 (#11)70.9 (#4)
Agent $/M$2.31$4.62
Input / output $/M$5.00 / $25.00$10.00 / $50.00
Context1M1.1M
ReleasedJul 24, 2026Sep 4, 2026
WeightsClosedClosed
API hosts66

Don’t pick one. Route.

Shogo can run Claude Opus 5 and GPT-6 Astra side by side and send each step of an agent to the one that’s cheaper for it.

Try both in Shogo