VerdictPal
Join

Models / OpenAI

GPT-5.6 Luna

GPT-5.6 · Released 2026-07-09

OpenAI's fastest and most cost-efficient generally available GPT-5.6 model, aimed at high-volume work and priced at $1 / $6 per 1M tokens.

GPT-5.6 tier for high-volume sub-agent, extraction, routing, and everyday coding workloads.

#6 of 51 on AA Index · current snapshot
51 AA Index · 9 benchmark rows · 1 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite59
ContextUnlisted
Input / 1M$1.0
Output / 1M$6.0
Knowledge cutoffunpublished
Output speed203 t/s
TTFT (default API)83.80s
Statussolid
ReasoningCodingAgenticLong context
Reasoning
91
Coding
53
Agentic
81
Long context
74

75average across 4 tested families

Editor's note
Luna moved from limited preview to general availability on July 9, 2026. Independent benchmark rows now cover the family, while OpenAI still has not published the context limit.
Benchmark placements

Where GPT-5.6 Luna places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
AA-LCRLong context#5/ 6474Artificial Analysis2026-07-21
AA time to first tokenPerformance#5/ 69Artificial Analysis2026-07-21
Terminal-BenchAgentic#9/ 6981Artificial Analysis2026-07-21
AA output speedPerformance#11/ 69Artificial Analysis2026-07-21
Artificial Analysis Intelligence IndexReasoning#12/ 6751Artificial Analysis2026-07-21
GPQA DiamondReasoning#17/ 6991Artificial Analysis2026-07-21
SciCodeCoding#20/ 6553Artificial Analysis2026-07-21
Humanity's Last ExamReasoning#22/ 6537Artificial Analysis2026-07-21
τ³-Bench BankingAgentic#31/ 6427Artificial Analysis2026-07-21
Family context

The three highest-scoring carded models in each capability family. Where GPT-5.6 Luna shows up, it's highlighted.

ReasoningGPT-5.6 Luna · 91
  1. 1
    Claude Mythos Preview
    95
  2. 2
    GPT-5.6 Sol
    94
  3. 3
    Kimi K3
    94
CodingGPT-5.6 Luna · 53
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticGPT-5.6 Luna · 81
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextGPT-5.6 Luna · 74
  1. 1
    Kimi K3
    75
  2. 2
    GPT-5.1
    75
  3. 3
    GPT-5.6 Sol
    74
Newest receipts

Benchmark rows added to the public ledger for GPT-5.6 Luna in the last 120 days. Older rows live in the full table below.

9 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context74%Artificial Analysis2
AA output speedPerformance203 t/sArtificial Analysis2
AA time to first tokenPerformance83.80sArtificial Analysis2
Artificial Analysis Intelligence IndexReasoning51.2%Artificial Analysis2
GPQA DiamondReasoning91.1%Artificial Analysis2
Humanity's Last ExamReasoning37.2%Artificial Analysis2
SciCodeCoding52.5%Artificial Analysis2
τ³-Bench BankingAgentic27.22%Artificial Analysis2
Terminal-BenchAgentic80.9%Artificial Analysis2
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Reasoning91
Coding53
Agentic81
Long context74
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover or tab any dot for name, score, and input price.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotMiniMaxxAIMetaZhipuXiaomiMeituanCohere

Full benchmark ledger

Every catalog benchmark for GPT-5.6 Luna. Scores link to the original source; gaps mean no public row exists yet.

9 sourced rows

9 of 32 catalog benchmarks have a sourced row for GPT-5.6 Luna.

Reasoning

Coding

Agentic

Long context

Performance

Changelog
  1. Recorded general availability across ChatGPT, Codex, and the API and refreshed benchmark evidence.
  2. Updated public pricing: Luna is listed at $1 input / $6 output per 1M tokens.
  3. Previewed as the fastest and most cost-efficient GPT-5.6 tier.
Sources