SShogo Bench

Claude Opus 4.7 vs GPT-6 Astra

GPT-6 Astra leads Claude Opus 4.7 by 16.3 points on the Shogo Agent Index. Claude Opus 4.7 wins 0 and GPT-6 Astra wins 22 of 22 shared benchmarks. Claude Opus 4.7 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarClaude Opus 4.7GPT-6 Astra
Agentic coding6080
Long-horizon autonomy6980
Knowledge-work agents4156
Computer use4467
Security & context learning5263

Shared benchmarks

BenchmarkClaude Opus 4.7GPT-6 Astra
APEX-Agentsindex49.2%64.7%
Vending-Bench 2index$10,937$15,515
GSO-Benchindex44.1%79.4%
PostTrainBenchindex24.3%41.9%
ARC-AGI93.5%98.5%
ARC-AGI-275.8%95.0%
Chess Puzzles41.0%76.0%
CritPt12.0%31.7%
DTBench94.7%97.3%
FrontierCode38.5%53.3%
FrontierMath Tier 431.7%97.6%
FrontierMath Tiers 1-370.2%93.7%
GPQA Diamond90.2%95.8%
Humanity's Last Exam36.2%54.8%
LMCA52.2%64.4%
MirrorCode43.3%60.0%
OTIS Mock AIME97.8%100.0%
ProofBench54.0%99.0%
SciCode54.5%56.5%
SimpleBench61.7%86.5%
SimpleQA Verified51.7%75.6%
WebDev Arena15671800

Specs and price

Claude Opus 4.7GPT-6 Astra
Agent Index54.6 (#29)70.9 (#4)
Agent $/M$2.31$4.62
Input / output $/M$5.00 / $25.00$10.00 / $50.00
Context1M1.1M
ReleasedApr 16, 2026Sep 4, 2026
WeightsClosedClosed
API hosts56

Don’t pick one. Route.

Shogo can run Claude Opus 4.7 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