SShogo Bench

Claude Opus 4.8 vs GPT-5.6 Luna

Claude Opus 4.8 leads GPT-5.6 Luna by 4.9 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Claude Opus 4.8 wins 14 and GPT-5.6 Luna wins 5 of 20 shared benchmarks. GPT-5.6 Luna is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarClaude Opus 4.8GPT-5.6 Luna
Agentic coding6359
Long-horizon autonomy6661
Knowledge-work agents4339
Computer use4740
Security & context learning5551

Shared benchmarks

BenchmarkClaude Opus 4.8GPT-5.6 Luna
APEX-Agentsindex48.9%43.0%
DeepSWEindex59.0%67.2%
Vending-Bench 2index$5,787$4,095
ARC-AGI92.5%88.0%
ARC-AGI-272.1%59.5%
Chess Puzzles52.0%43.0%
CritPt20.9%20.6%
DTBench94.9%89.1%
FrontierCode46.5%39.8%
FrontierMath Tier 456.1%61.0%
FrontierMath Tiers 1-380.0%82.1%
GPQA Diamond91.0%91.6%
LMCA57.5%48.5%
MirrorCode65.6%21.1%
OTIS Mock AIME98.3%98.3%
ProofBench69.0%60.0%
SciCode53.5%53.6%
SimpleBench64.8%46.8%
SimpleQA Verified53.0%41.0%
WebDev Arena15591519

Specs and price

Claude Opus 4.8GPT-5.6 Luna
Agent Index56.3 (#20)51.4 (#34)
Agent $/M$2.31$0.10
Input / output $/M$5.00 / $25.00$0.20 / $1.20
Context1M1.1M
ReleasedMay 28, 2026Jul 9, 2026
WeightsClosedClosed
API hosts65

Don’t pick one. Route.

Shogo can run Claude Opus 4.8 and GPT-5.6 Luna side by side and send each step of an agent to the one that’s cheaper for it.

Try both in Shogo