SShogo Bench

Claude Fable 5.2 vs GPT-5.4

Claude Fable 5.2 leads GPT-5.4 by 24.8 points on the Shogo Agent Index. Claude Fable 5.2 wins 16 and GPT-5.4 wins 0 of 16 shared benchmarks. GPT-5.4 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarClaude Fable 5.2GPT-5.4
Agentic coding8657
Long-horizon autonomy7861
Knowledge-work agents6740
Computer use7742
Security & context learning6554

Shared benchmarks

BenchmarkClaude Fable 5.2GPT-5.4
APEX-Agentsindex80.1%52.4%
Vending-Bench 2index$9,628$6,144
PostTrainBenchindex44.6%17.8%
ARC-AGI98.5%93.7%
ARC-AGI-293.8%74.0%
Chess Puzzles79.0%48.0%
CritPt35.7%23.4%
DTBench99.2%94.4%
FrontierMath Tier 4100.0%49.0%
FrontierMath Tiers 1-395.1%78.6%
GPQA Diamond96.0%93.3%
LMCA70.6%52.0%
OTIS Mock AIME99.7%97.8%
ProofBench100.0%56.0%
SciCode67.7%56.6%
SimpleQA Verified94.0%45.1%

Specs and price

Claude Fable 5.2GPT-5.4
Agent Index76.4 (#1)51.6 (#33)
Agent $/M$3.21$1.27
Input / output $/M$8.00 / $40.00$2.50 / $15.00
Context1M1.1M
ReleasedNov 9, 2026Mar 5, 2026
WeightsClosedClosed
API hosts44

Don’t pick one. Route.

Shogo can run Claude Fable 5.2 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