Request a pilot
‹ PublicAI Index

The LLM benchmark aggregator.

GPT-6 Luna

OpenAI

Strongest in Routing & classification (#1 of 85), weakest in Toxicity avoidance (#210 of 272). Above par in 19 of 29 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5.5 and Claude Opus 5 and ahead of Gemini 3.5 Flash and GLM-5.1.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Decisions72.7#2/85
Routing & classification74#1/85
Calibration71.3#2/85
Coding55.5#30/165
Agentic coding56.3#32/157
Code generation53.3#34/77
Agents59.5#35/268
Knowledge work62.4#29/178
Human preference61#74/342
Human preference61#74/342
Reasoning51.3#80/178
Reasoning50.8#58/140
Mathematics51.8#72/140
Professional49.7#82/168
Biology research36.5#14/16
Medical54.3#48/140
Finance52.1#76/151
Legal48.5#90/151
Safety53.4#88/337
Secure code63.9#21/274
Fairness60.2#42/300
Harm refusal49.8#159/300
Jailbreak resistance50#181/272
Toxicity avoidance48.9#210/272
Core abilities44.5#154/204
Data analysis50#35/57
Language39.7#46/57
Instruction following28.9#55/57
General intelligence52#80/204

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for GPT-6 Luna, left for the other.

Grok Build 0.1won 18 · lost 1
DeepSeek V3won 20 · lost 2
GPT-4o Miniwon 18 · lost 3
GPT-4.1 Nanowon 18 · lost 3
GPT-5.4 Nanowon 17 · lost 3
Kimi K2won 17 · lost 3
Llama 4 Scout Instructwon 16 · lost 3
Command Awon 16 · lost 3
Grok 4.3won 21 · lost 4
GPT OSS 20Bwon 15 · lost 3
Gemini 2.5 Flashwon 17 · lost 4
DeepSeek V3.2won 17 · lost 4
Mimo V2.5 Prowon 16 · lost 4
Kimi K2.7 Codewon 15 · lost 4
GPT-5.4 Miniwon 15 · lost 4
Gemini 3.5 Flash Litewon 15 · lost 4
Gemini 3.1 Flash Litewon 14 · lost 4
Gemini 2.0 Flashwon 14 · lost 5
Qwen3.6 Pluswon 14 · lost 5
GPT-4owon 13 · lost 5
Grok 4.1 Fastwon 13 · lost 5
GPT OSS 120Bwon 15 · lost 6
Claude Haiku 4.5won 15 · lost 6
Gemma 4 31Bwon 15 · lost 6
DeepSeek R1won 15 · lost 7
Kimi K2.6won 17 · lost 8
GPT-4.1 Miniwon 14 · lost 7
O3 Miniwon 13 · lost 7
GLM-4.6won 13 · lost 7
MiniMax M3won 16 · lost 9
Kimi K2.5won 13 · lost 8
Claude 3.7 Sonnetwon 11 · lost 7
GPT-4.1won 12 · lost 8
Claude Sonnet 4won 13 · lost 9
Qwen3.7 Maxwon 14 · lost 11
GLM-5.2won 14 · lost 11
MiniMax M2.7won 10 · lost 8
Grok 4won 11 · lost 9
Qwen3 235B A22B Instructwon 11 · lost 9
DeepSeek V4 Flashwon 14 · lost 12
Gemini 3.6 Flashwon 14 · lost 12
GLM-5.1won 10 · lost 9
Gemini 3.5 Flashwon 13 · lost 12
GPT-5 Nanowon 9 · lost 9
Inklingwon 12 · lost 13
Qwen3.8 27Bwon 9 · lost 10
Gemini 3.8 Flashwon 12 · lost 14
Claude Opus 4won 9 · lost 11
Gemini 2.5 Prowon 9 · lost 13
GLM-5.3 Flashwon 10 · lost 15
O4 Miniwon 8 · lost 13
GPT-5.6 Lunawon 11 · lost 18
Gemini 3.1 Prowon 7 · lost 13
Claude Sonnet 4.6won 8 · lost 15
GPT-5 Miniwon 7 · lost 14
GPT-5won 7 · lost 14
Claude Opus 4.6won 7 · lost 14
Grok 4.5won 8 · lost 17
GPT-5.4won 8 · lost 17
Muse Spark 1.1won 8 · lost 17
GPT-5.1won 6 · lost 13
Gemini 3.7 Flashwon 6 · lost 13
GPT-5.2won 7 · lost 17
O3won 6 · lost 15
Claude Sonnet 4.5won 6 · lost 15
DeepSeek V4 Prowon 6 · lost 19
Grok 4.7won 6 · lost 20
GPT-6 Solwon 6 · lost 20
Claude Opus 4.5won 4 · lost 15
Qwen3.8 Maxwon 5 · lost 20
Claude Opus 4.7won 5 · lost 20
Mimo V2.6 Prowon 3 · lost 16
GPT-5.6 Solwon 4 · lost 22
Claude Sonnet 5won 3 · lost 22
GPT-5.6 Terrawon 3 · lost 22
GPT-5.5won 3 · lost 22
GLM-5.3won 2 · lost 17
Muse Spark 1.2won 2 · lost 18
DeepSeek V4.1 Flashwon 2 · lost 21
Claude Opus 4.8won 2 · lost 23
Kimi K3won 1 · lost 19
GPT-6 Astrawon 1 · lost 20
Grok 4.6won 1 · lost 25
Muse Spark 1.3won 0 · lost 18
Claude Fable 5.1won 0 · lost 19
Claude Fable 5won 0 · lost 19
Claude Opus 5won 0 · lost 26
Claude Opus 5.5won 0 · lost 26

§ 3 · Sources

Where the numbers come from

9 publications, 30 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-26
LMArena Text 1442
Artificial Analysis ↗1 measureread 2026-09-26
Artificial Analysis Intelligence Index 37
GDPval-AA ↗1 measureread 2026-09-26
GDPval-AA 1367
AA-Briefcase ↗1 measureread 2026-09-26
AA-Briefcase 1299
ARC-AGI-2 ↗1 measureread 2026-09-26
ARC-AGI-2 59.3%
LiveBench ↗8 measuresread 2026-09-26
LiveBench 72LiveBench · Reasoning 81.8LiveBench · Coding 79LiveBench · Agentic Coding 51.2LiveBench · Mathematics 89.1LiveBench · Data Analysis 73.4LiveBench · Language 73.8LiveBench · Instruction Following 55.9
Vals.ai ↗9 measuresread 2026-09-26
Vals · Legal Research Bench 30.29%Vals · Harvey Legal Agent Benchmark 2.92%Vals · Finance Agent 49.87%Vals · MedCode 44.69%Vals · MedScribe 83.71%Vals · BioMysteryBench 61.48%Vals · Vibe Code Bench 81.65% (OpenHands)Vals · Code Migration 42.55%Vals · ProofBench 64%
JevBench ↗2 measuresread 2026-09-26
JevBench · Intelligence 97.4%JevBench · Calibration 93.5%
Enkrypt AI Safety Leaderboard ↗6 measuresread 2026-09-26
Enkrypt · Jailbreak risk 12.8%Enkrypt · Harmful content risk 0%Enkrypt · CBRN risk 33%Enkrypt · Toxicity risk 7.1%Enkrypt · Bias risk 62%Enkrypt · Insecure code risk 5.3%

Badge

PublicAI Index badge for GPT-6 Luna[![PublicAI Index](https://publicai.io/model-index/badge?model=gpt-6-luna)](https://publicai.io/model-index/m/gpt-6-luna)