SShogo Bench

Claude Fable 5.2 vs Grok 4.6

Claude Fable 5.2 leads Grok 4.6 by 20.2 points on the Shogo Agent Index. Claude Fable 5.2 wins 18 and Grok 4.6 wins 0 of 18 shared benchmarks. Grok 4.6 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarClaude Fable 5.2Grok 4.6
Agentic coding8661
Long-horizon autonomy7868
Knowledge-work agents6747
Computer use7742
Security & context learning6552

Shared benchmarks

BenchmarkClaude Fable 5.2Grok 4.6
APEX-Agentsindex80.1%65.3%
Vending-Bench 2index$9,628$9,047
FrontierSWEindex69.9%25.3%
CursorBenchindex60.0%41.4%
ARC-AGI98.5%87.5%
ARC-AGI-293.8%67.1%
Chess Puzzles79.0%33.0%
CritPt35.7%19.7%
DTBench99.2%97.3%
FrontierMath Tier 4100.0%31.7%
FrontierMath Tiers 1-395.1%66.0%
GPQA Diamond96.0%94.0%
LMCA70.6%48.5%
OTIS Mock AIME99.7%99.2%
ProofBench100.0%51.0%
SciCode67.7%56.5%
SimpleBench94.9%75.9%
SimpleQA Verified94.0%49.3%

Specs and price

Claude Fable 5.2Grok 4.6
Agent Index76.4 (#1)56.2 (#22)
Agent $/M$3.21$0.98
Input / output $/M$8.00 / $40.00$2.00 / $6.00
Context1M500K
ReleasedNov 9, 2026Aug 12, 2026
WeightsClosedClosed
API hosts44

Don’t pick one. Route.

Shogo can run Claude Fable 5.2 and Grok 4.6 side by side and send each step of an agent to the one that’s cheaper for it.

Try both in Shogo