SShogo Bench

Claude Opus 4.8 vs GPT-5.4

Claude Opus 4.8 leads GPT-5.4 by 4.7 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Claude Opus 4.8 wins 14 and GPT-5.4 wins 7 of 21 shared benchmarks. GPT-5.4 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarClaude Opus 4.8GPT-5.4
Agentic coding6357
Long-horizon autonomy6661
Knowledge-work agents4340
Computer use4742
Security & context learning5554

Shared benchmarks

BenchmarkClaude Opus 4.8GPT-5.4
APEX-Agentsindex48.9%52.4%
DeepSWEindex59.0%51.8%
Vending-Bench 2index$5,787$6,144
DeepResearch Benchindex50.2%35.1%
GSO-Benchindex47.1%31.4%
PostTrainBenchindex31.0%17.8%
ARC-AGI92.5%93.7%
ARC-AGI-272.1%74.0%
Chess Puzzles52.0%48.0%
CritPt20.9%23.4%
DTBench94.9%94.4%
FrontierMath Tier 456.1%49.0%
FrontierMath Tiers 1-380.0%78.6%
GPQA Diamond91.0%93.3%
LMCA57.5%52.0%
MirrorCode65.6%23.3%
OTIS Mock AIME98.3%97.8%
ProofBench69.0%56.0%
SciCode53.5%56.6%
SimpleQA Verified53.0%45.1%
WebDev Arena15591463

Specs and price

Claude Opus 4.8GPT-5.4
Agent Index56.3 (#20)51.6 (#33)
Agent $/M$2.31$1.27
Input / output $/M$5.00 / $25.00$2.50 / $15.00
Context1M1.1M
ReleasedMay 28, 2026Mar 5, 2026
WeightsClosedClosed
API hosts64

Don’t pick one. Route.

Shogo can run Claude Opus 4.8 and GPT-5.4 side by side and send each step of an agent to the one that’s cheaper for it.

Try both in Shogo