Skip to content
Leaderboard / OpenAI

GPT-6 Luna

OpenAI Max CLOSED Proprietary
Other settings:MediumLow
Verdict

GPT-6 Luna is a solid mid-pack model: #90 of 406 by SIscope score. Strongest at math, weaker at chat quality. Cheaper than 88% of priced models.

SIscope score
69
#90 of 406
Human-vote rank
#170
of 419 · score 1389
Price per 1M tokens, blended
$0.2
$0.1 in · $0.5 out
Context window
1.1M
1,050,000 tokens

Shape

Each spoke is a category, 0 at the centre, 100 at the edge.

CodingAgentsReasoningMathChatInstructions

Category scores

Percentiles: 90 means better than about 90% of tested models.

Coding72
Agents & tools63
Reasoning & knowledge67
Math80
Chat quality60
Following instructions73

What goes into each category · Compare with DeepSeek V4.1 Flash

Human-vote score over time

Text arena, recent updates.

138513901395Oct 20: 1389.4 (rank #167)Oct 22: 1390.3 (rank #168)Oct 28: 1389.2 (rank #170)Oct 20Oct 28

Coding votes over time

WebDev arena, recent updates.

157015781585Oct 16: 1578.2 (rank #35)Oct 18: 1576.6 (rank #36)Oct 19: 1577.8 (rank #38)Oct 22: 1578.4 (rank #38)Oct 23: 1578.6 (rank #39)Oct 27: 1579.8 (rank #40)Oct 28: 1579.4 (rank #41)Oct 16Oct 28

Independent test runs

Run by Epoch AI on the same harness for every model. Bars show the share of questions answered correctly.

GPQA Diamond90%
PhD-level science multiple choice questions
Better than 86% of 333 tested models. Run Sep 22, 2026.
SimpleQA Verified41%
Short factual questions; rewards not making things up
Better than 52% of 100 tested models. Run Sep 22, 2026.
OTIS Mock AIME 2024-2599%
Competition maths problems in the style of AIME
Better than 92% of 313 tested models. Run Sep 22, 2026.
FrontierMath Tiers 1-379%
Very hard, unpublished research-style maths problems
Better than 85% of 122 tested models. Run Sep 22, 2026.

Human-vote arenas

People compare two anonymous answers and pick the better one. Scores work like chess ratings.

ArenaRankScoreRange (95%)VotesUpdated
Text arena
Everyday chat, blind human votes
#170 of 41913891385–139386,478Oct 28, 2026
WebDev arena
Building web apps, blind human votes
#41 of 15215791574–158527,808Oct 28, 2026
Vision arena
Questions about images, blind human votes
#85 of 16712161205–12273,662Oct 19, 2026
Agent arena
Real multi-step agent sessions
#22 of 571.9%1.5–2.3%223,505Nov 2, 2026

Source: Arena (formerly LMArena) leaderboard dataset, CC BY 4.0. Fetched Nov 4, 2026.

Price

Input
$0.1
per 1M tokens
Output
$0.5
per 1M tokens
Blended
$0.2
3 in : 1 out, per 1M
Context window
1.1M
max tokens per request
Speed
Coming soon
our own tests

List price from the OpenRouter models API (openai/gpt-6-luna), Nov 4, 2026. Providers can charge differently.