01Overview02Models03Rankings04Compare05Quiz06Guide

Model

GPT-5.6 Luna NO.09

The fast, cost-efficient GPT-5.6 tier for everyday work at scale.

CompareBack to the index

OpenAI · GPT · Fast · Closed weights

Specification

Fig.01 — GPT — OpenAI

USA

Context window
1.1Mtokens
Max output
128Ktokens
Input price
$1/ 1M
Output price
$6/ 1M
Cached input
$0.10/ 1M
Throughput
—
Class
Fast
Modalities
Text, Image, Pdf
Weights
Closed
Knowledge cutoff
Feb 2026
Released
Jul 2026

Overview

GPT-5.6 Luna is the fastest and most cost-efficient member of the GPT-5.6 family, built for responsive everyday work — summarization, drafting, and routine automation — where throughput and cost matter more than maximum depth. Despite its low $1/$6 pricing it is remarkably capable: on Artificial Analysis' Coding Agent Index (74.6) it edges Claude Opus 4.8 at roughly a quarter of the cost. It keeps the family's ~1.05M-token context, 128K output, and February-2026 knowledge, and replaces the old mini/nano split with a single strong, cheap model.

Strengths

  • Lowest GPT-5.6 price ($1/$6) with strong quality
  • Surprising agentic-coding value (Coding Agent Index 74.6)
  • Fast, low-latency responses at scale
  • ~1.05M-token context with multimodal input
  • Cheap, predictable prompt caching

Trade-offs

  • Not a substitute for Sol on hard long-horizon coding
  • Pricier than the older GPT-5.4 nano/mini for bulk work
  • No native audio or video input

Fit

Best for

  • High-volume summarization and drafting
  • Routine automation and everyday assistants
  • Cost-sensitive coding and tool use
  • Latency-sensitive production traffic

Not ideal for

  • Hardest autonomous coding runs (use Sol)
  • Rock-bottom bulk classification (GPT-5.4 nano is cheaper)

Capabilities

Ten axes, normalized 0–100 and scored the same way across the whole catalog — so a 78 here means what a 78 means anywhere else on the bench.

Fig.02 — Capability radar

0 — 100

Reasoning
86
Coding
86
Math
85
Writing
85
Knowledge
86
Speed
86
Agentic
84
Vision
84
Multilingual
87
Long Context
88

Benchmarks

Public results, with independent third-party runs marked. Bars normalize percentages against 100 and Elo ratings against a 1500 ceiling.

Independent index

Artificial Analysis Intelligence Index

Composite of ~9–10 independent evals · Artificial Analysis

51/ 100

BenchmarkResult
Artificial Analysis Intelligence IndexIndependentGeneral · Artificial Analysis
51%
SWE-bench ProIndependentCoding · SWE-bench Pro
62.7%

Alternatives

Models in roughly the same class — the ones worth weighing against this record.

Compare side by side
01GPT-5.6 TerraOpenAIThe GPT-5.6 balanced workhorse — GPT-5.5-class quality at half the price.02Claude Sonnet 4.6AnthropicThe production workhorse — best balance of speed and intelligence.03Gemini 3.5 FlashGoogle DeepMindFast, cheap multimodal workhorse with near-Pro quality.

Index

PreviousGPT-5.6 TerraNextGPT-5.6 Sol Ultra
All models/EDUCATION