The Model Index
Up to four models

GPT-6 Luna vs Gemini 3.1 Pro

On the index (GPT-4 = 100), GPT-6 Luna scores 162 and Gemini 3.1 Pro 161. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.

Next to GPT-6 Luna
Presets
GPT-6 Luna
162146–178
WeightsClosed
ReleasedSep 2026
Price in / out$0.10 / $0.50
Context1.05M
Core tests taken6 of 9
Gemini 3.1 Pro
161146–177
WeightsClosed
ReleasedFeb 2026
Price in / out$2.0 / $12
Context1.05M
Core tests taken9 of 9

GPT-6 Luna and Gemini 3.1 Pro are too close to call (1 point apart, within the uncertainty), ahead on 4 of 6 shared core tests, and costs 23× less.

Test by testthird-party results

GPT-6 LunaGemini 3.1 Pro
0%25%50%75%100%REASONING & KNOWLEDGEGPQA Diamond (Epoch AI)GPQA Diamond (Epoch AI)SimpleQA VerifiedSimpleQA VerifiedMATHSFrontierMath (tiers 1-3)FrontierMath (tiers 1-3)AIME-style maths (OTIS mock)AIME-style maths (OTIS mock)ProofBench v1.1ProofBench v1.1Chess puzzlesChess puzzlesFrontierMath (tier 4)FrontierMath (tier 4)CODINGTerminal-Bench 4.0Terminal-Bench 4.0Code MigrationCode MigrationIOIIOIProgramBenchProgramBenchVibe Code Bench v1.1Vibe Code Bench v1.1MirrorCodeMirrorCodeAGENTSAPEX-AgentsAPEX-AgentsBioMysteryBenchBioMysteryBenchSRE BenchSRE BenchTerminal-Bench-ScienceTerminal-Bench-ScienceDTBenchDTBenchFurniture assemblyFurniture assemblyLMCALMCAMystery game puzzlesMystery game puzzlesPROFESSIONAL WORKFinance Agent (v2)Finance Agent (v2)EMBEMBLegal Research BenchLegal Research BenchMedCodeMedCodeMedScribeMedScribePublic Benefits Bench v1.1Public Benefits Bench v1.1SAGESAGETax Agent BenchTax Agent BenchNOVEL PROBLEMSARC-AGI-2ARC-AGI-2ARC-AGI-1ARC-AGI-1ARC-AGI-3ARC-AGI-3MysteryMechanismMysteryMechanism
Every result as a table
TestGPT-6 LunaGemini 3.1 Pro
Reasoning & knowledge
GPQA Diamond (Epoch AI)90.5%94.4%
SimpleQA Verified41.4%73.5%
Maths
FrontierMath (tiers 1-3)79.0%59.7%
AIME-style maths (OTIS mock)98.9%95.6%
ProofBench v1.164.0%26.0%
Chess puzzles36.0%51.0%
FrontierMath (tier 4)56.1%26.8%
Coding
Terminal-Bench 4.013.6%2.5%
Code Migration42.5%17.3%
IOI55.6%51.8%
ProgramBench0.5%0.0%
Vibe Code Bench v1.181.7%32.0%
MirrorCode26.7%8.9%
Agents
APEX-Agents44.3%35.1%
BioMysteryBench61.5%60.0%
SRE Bench2.7%0.0%
Terminal-Bench-Science4.3%1.4%
DTBench90.1%97.1%
Furniture assembly44.2%26.7%
LMCA44.5%53.8%
Mystery game puzzles7.0%34.0%
Professional work
Finance Agent (v2)49.9%43.0%
EMB68.5%52.6%
Legal Research Bench30.3%20.7%
MedCode44.7%59.1%
MedScribe83.7%76.1%
Public Benefits Bench v1.157.6%53.8%
SAGE83.4%83.2%
Tax Agent Bench20.0%9.3%
Novel problems
ARC-AGI-259.3%77.1%
ARC-AGI-186.7%98.0%
ARC-AGI-30.2%0.4%
MysteryMechanism19.4%20.7%
Task length · METR46 sec

Tests at least two of these models have taken. Bold rows are the core tests.

This page as a chartDownload PNG