SShogo Bench

Gemini 3.7 Flash vs GPT-5.6 Luna

Gemini 3.7 Flash leads GPT-5.6 Luna by 4.3 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Gemini 3.7 Flash wins 12 and GPT-5.6 Luna wins 6 of 18 shared benchmarks. GPT-5.6 Luna is cheaper for agent workloads.

Add more models to this comparison →

Pillars

PillarGemini 3.7 FlashGPT-5.6 Luna
Agentic coding6159
Long-horizon autonomy6461
Knowledge-work agents4839
Computer use4340
Security & context learning5351

Shared benchmarks

BenchmarkGemini 3.7 FlashGPT-5.6 Luna
APEX-Agentsindex67.8%43.0%
DeepSWEindex65.5%67.2%
ARC-AGI95.5%88.0%
ARC-AGI-284.6%59.5%
Chess Puzzles66.0%43.0%
CritPt14.3%20.6%
DTBench96.8%89.1%
FrontierCode43.6%39.8%
FrontierMath Tier 436.6%61.0%
FrontierMath Tiers 1-371.6%82.1%
GPQA Diamond94.8%91.6%
LMCA50.4%48.5%
MirrorCode41.1%21.1%
OTIS Mock AIME97.2%98.3%
ProofBench58.0%60.0%
SciCode59.8%53.6%
SimpleQA Verified69.2%41.0%
WebDev Arena15871519

Specs and price

Gemini 3.7 FlashGPT-5.6 Luna
Agent Index55.7 (#23)51.4 (#34)
Agent $/M$0.35$0.10
Input / output $/M$0.75 / $3.75$0.20 / $1.20
Context1M1.1M
ReleasedAug 13, 2026Jul 9, 2026
WeightsClosedClosed
API hosts65

Don’t pick one. Route.

Shogo can run Gemini 3.7 Flash 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