SShogo Bench

GPT-6.1 Sol vs Grok 4.5

GPT-6.1 Sol leads Grok 4.5 by 20.6 points on the Shogo Agent Index. GPT-6.1 Sol wins 19 and Grok 4.5 wins 0 of 20 shared benchmarks. Grok 4.5 is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarGPT-6.1 SolGrok 4.5
Agentic coding8054
Long-horizon autonomy7859
Knowledge-work agents5442
Computer use6636
Security & context learning6248

Shared benchmarks

BenchmarkGPT-6.1 SolGrok 4.5
APEX-Agentsindex60.0%56.2%
Vending-Bench 2index$12,861$3,887
PostTrainBenchindex38.8%22.6%
ARC-AGI98.5%87.2%
ARC-AGI-294.2%52.6%
Chess Puzzles68.0%35.0%
CritPt31.7%15.4%
DTBench96.5%96.5%
FrontierCode50.2%42.4%
FrontierMath Tier 4100.0%24.4%
FrontierMath Tiers 1-393.7%57.2%
GPQA Diamond95.4%93.4%
LMCA62.8%45.2%
MirrorCode61.1%32.2%
OTIS Mock AIME100.0%97.8%
ProofBench99.0%31.0%
SciCode55.8%55.0%
SimpleBench82.9%70.0%
SimpleQA Verified73.9%48.3%
WebDev Arena17591555

Specs and price

GPT-6.1 SolGrok 4.5
Agent Index70 (#5)49.4 (#36)
Agent $/M$0.84$0.81
Input / output $/M$2.00 / $10.00$2.00 / $6.00
Context1.1M500K
ReleasedSep 29, 2026Jul 8, 2026
WeightsClosedClosed
API hosts64

Don’t pick one. Route.

Shogo can run GPT-6.1 Sol and Grok 4.5 side by side and send each step of an agent to the one that’s cheaper for it.

Try both in Shogo