Claude Fable 5$22.000/MClaude Opus 5$11.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.6 Sol$12.500/MGPT-5.6 Terra$5.000/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/MClaude Fable 5$22.000/MClaude Opus 5$11.000/MClaude Opus 4.8$11.000/MClaude Opus 4.7$11.000/MClaude Opus 4.6$11.000/MClaude Opus 4.5$33.000/MClaude Sonnet 3.7$6.600/MClaude Opus 3$33.000/MClaude 2.1$12.800/MClaude 2$12.800/MGPT-5.6 Sol$12.500/MGPT-5.6 Terra$5.000/MGPT-5.5$12.500/MGPT-5.2$5.425/MGPT-5.2-Codex$5.425/MGPT-5$3.875/MGPT-4.5$97.500/MGPT-4 Turbo Preview$16.000/MGPT-4$39.000/MGPT-4-32k$78.000/Mo3$19.000/Mo3-mini$2.090/Mo4-mini$2.090/Mo1$28.500/Mo1-mini$5.700/Mo1-preview$28.500/MGemini 3.5 Pro$5.000/MGemini 3.1 Pro$5.000/MGemini 3 Pro$5.000/MGemini 2.5 Pro$3.875/M
OpenAI
OpenAI
Efficient

GPT-5.6 Luna

Efficient

The fastest and most affordable GPT-5.6 tier, a deliberately low-reasoning lane for high-volume work where speed and cost are the binding constraints rather than depth. It keeps the family's 1M-token context window at $0.20/$1.20 per MTok, which is 25× cheaper than Sol on input, with cached input at $0.02.

GPT-5.6 Luna is a efficient AI model from OpenAI. It costs $0.200 per million input tokens and $1.200 per million output tokens (blended $0.500/M), with a 1,050,000-token context window.

INPUT
$0.200/M
per million input tokens
OUTPUT
$1.200/M
per million output tokens
BLENDED 70/30
$0.500/M
unchanged since 3 May
CONTEXT
1,050,000
tokens
What it is good at
  • Lowest cost in the GPT-5.6 family, 25× cheaper input than Sol
  • Fast responses
  • High-volume throughput
  • 1M-token context
  • Cached input at $0.02 per MTok
Typical use cases
  • Classification and extraction at scale
  • Routing and triage
  • High-volume summarisation
  • Cheap intermediate agent steps

Benchmarks

vs. best public score
Graduate-level science questions, "Google-proof".
LMArena Elo1451 Elo
Crowd-sourced head-to-head preference Elo rating.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

Also reported by the provider

Results on benchmarks outside the set we compare across models. They are listed for completeness and are deliberately excluded from the bars above and from every ranking on this site, because a score is only comparable against the same benchmark variant. Why benchmark scores do not compare →

  • SWE-bench Pro62.7%self-reported by OpenAI · 2026-07 · SWE-bench Verified was not published for GPT-5.6.
  • Terminal-Bench 2.184.7%self-reported by OpenAI · 2026-07
  • FrontierMath Tier 1–3 (v2)78.6%self-reported by OpenAI · 2026-07
  • FrontierMath Tier 4 (v2)58.5%self-reported by OpenAI · 2026-07
  • MMMU Pro (no tools)78.4%self-reported by OpenAI · 2026-07
  • BrowseComp83.3%self-reported by OpenAI · 2026-07
  • OSWorld 2.045.6%self-reported by OpenAI · 2026-07
  • Agents' Last Exam50.3%self-reported by OpenAI · 2026-07
  • ARC-AGI-30.18%self-reported by OpenAI · 2026-07

More from OpenAI

See all 28

Frequently asked questions

How much does GPT-5.6 Luna cost?

GPT-5.6 Luna costs $0.200 per million input tokens and $1.200 per million output tokens, for a blended reference rate of $0.500 per million tokens.

What is GPT-5.6 Luna's context window?

GPT-5.6 Luna supports up to 1,050,000 tokens of context in a single request.

What is GPT-5.6 Luna best for?

GPT-5.6 Luna is well suited to Lowest cost in the GPT-5.6 family, 25× cheaper input than Sol, Fast responses and High-volume throughput.

Who makes GPT-5.6 Luna?

GPT-5.6 Luna is developed and served by OpenAI. It was released in Jul 2026.

Compared with

Terms used on this page