HolyGrAIL

All tasks / Models

GPT-6 Luna

OpenAI's most efficient GPT tier for focused, cost-sensitive, high-volume workloads.

  • OpenAI
  • Closed
  • Released 22 Sep 2026

Official page

Where it ranks

  • A
    Chat & assistants
    Models · Cloud · 1,443 on LMArena Text Arena (overall, style control)
  • A
    Search & research
    Models · Cloud · 1,209 on LMArena Search Arena (style control)
  • B
    Writing
    Models · Cloud · 1,410 on LMArena Text Arena: Creative Writing (style control)
  • –
    Documents & OCR
    Models · Cloud · no result on this task’s benchmarks yet
  • C
    Coding
    Models · Cloud · 1,577 on LMArena WebDev / Code Arena (overall)
  • A
    Image understanding
    Models · Cloud · 1,241 on LMArena Vision Arena (overall, style control)
  • C
    Vector graphics & logos
    Models · Cloud · 1,249 on Design Arena SVG
  • C
    Workflow automation
    Models · Cloud · 53.2% on AutomationBench-AA (Artificial Analysis)

Facts

Vendor
OpenAI
Family
GPT Luna
Released
22 Sep 2026
Weights
Closed: API and apps only
License
proprietary
API price
$0.1 in / $0.5 out per 1M tokens
Access
Paid only
Last verified
18 Oct 2026

Price from developers.openai.com, 18 Oct 2026.

Benchmark results

BenchmarkResultEntry on the leaderboardAs ofSource
Artificial Analysis Intelligence Index38GPT-6 Luna (Max)
Raw value from the page's embedded model data; the site displays it rounded to an integer.
18 Oct 2026artificialanalysis.ai
Artificial Analysis MMMU-Pro (independent evaluation)79.7%GPT-6 Luna (Xhigh)
Raw fraction 0.7971 from the model data embedded in https://artificialanalysis.ai/leaderboards/models, converted to percent; rank 28 of 124 entries with an MMMU-Pro result; no last-updated date shown.
19 Oct 2026artificialanalysis.ai
AutomationBench-AA (Artificial Analysis)53.2%GPT-6 Luna (Max)
AutomationBench-AA Score (share of each task's objectives completed, a task with any guardrail violation scoring zero); raw value from the page's embedded model data (displayed rounded); rank 24 of 32; Tasks Completed 20.24%, Objectives Completed 82.02%; no last-updated date shown, so as_of = retrieval date.
19 Oct 2026artificialanalysis.ai
Design Arena SVG1,249GPT-6 Luna
Design Arena SVG, rank 24, 610 battles; rows read from POST https://www.designarena.ai/api/leaderboard (arenaType 'models', category 'svg'); names from https://www.designarena.ai/api/registry
19 Oct 2026designarena.ai
LMArena Search Arena (style control)1,209gpt-6-luna-max
From LMArena's official leaderboard-dataset (search_style_control, split latest, category overall, published 2026-10-13), which mirrors https://arena.ai/leaderboard/search (site blocks automated fetches). Rank 13 of 42; 10793 votes; 95% CI 1201.9-1216.6.
13 Oct 2026huggingface.co
LMArena Text Arena (overall, style control)1,443gpt-6-luna-max
From LMArena's official leaderboard-dataset (text_style_control, split latest, category overall, published 2026-10-15), which mirrors https://arena.ai/leaderboard/text (site blocks automated fetches).
15 Oct 2026huggingface.co
LMArena Text Arena: Creative Writing (style control)1,410gpt-6-luna-max
From LMArena's official leaderboard-dataset (text_style_control, split latest, category creative_writing, published 2026-10-15), which mirrors https://arena.ai/leaderboard/text/creative-writing (site blocks automated fetches). Rank 91 of 414; 11078 votes; 95% CI 1402.9-1417.2.
15 Oct 2026huggingface.co
LMArena Vision Arena (overall, style control)1,241gpt-6-luna-max
From LMArena's official leaderboard-dataset (vision_style_control, split latest, category overall, published 2026-10-09), which mirrors https://arena.ai/leaderboard/vision (site blocks automated fetches). Rank 65 of 165; 2833 votes; 95% CI 1228.8-1253.1.
9 Oct 2026huggingface.co
LMArena WebDev / Code Arena (overall)1,577GPT-6 Luna (Max)
From LMArena's official leaderboard-dataset (webdev = Code Arena, split latest, category overall, published 2026-10-18), which mirrors https://arena.ai/leaderboard/webdev. Rank 36; 20845 votes; 95% CI 1570.4-1582.8.
18 Oct 2026huggingface.co
ParseBench (models)65.8%OpenAI GPT-6 Luna (Reasoning Max)
Overall (mean of tables, charts, content faithfulness, semantic formatting, visual grounding) from leaderboard.csv in the ParseBench repo (last leaderboard commit 2026-10-09). Best of six reasoning levels.
9 Oct 2026github.com

Where you can use it

Earlier versions

  • GPT-5.6 Luna · 9 Jul 2026

Sources