GPT-5.6 Luna

GPT-5.6 Luna is OpenAI's generally available GPT-5.6 tier for cost-sensitive, high-volume reasoning, coding triage, and production workloads.

Platform: OpenAI

Language ModelReasoningCodingOpenAI

OpenAI-hosted

$0.20/1M input, $0.02/1M cached input, and $1.20/1M output tokens; prompts above 272K input use 2x input and 1.5x output rates; cache writes cost 1.25x uncached input

Commercial

🚀Function Overview

The most affordable GPT-5.6 tier for high-volume work, now priced 80% below its July 9 launch rates for input, cached input, and output tokens.

Key Features

Official OpenAI GPT-5.6 family member with model ID gpt-5.6-luna
OpenAI's most cost-efficient GPT-5.6 tier, roughly corresponding to earlier nano tiers
Generally available across the OpenAI API, ChatGPT, and Codex
1,050,000-token context window and 128,000-token maximum output
February 16, 2026 knowledge cutoff with reasoning-token support
$0.20 input, $0.02 cached input, and $1.20 output per 1 million tokens after an 80% reduction from the July 9 launch rates
Prompts above 272K input tokens use 2x input and 1.5x output rates; cache writes use 1.25x uncached input

Use Cases

•Low-cost GPT-5.6 family smoke tests
•Coding triage before escalation to Terra or Sol
•Model-routing experiments where price is the first constraint
•Reader comparison against older low-cost OpenAI model choices
•Re-evaluating production routes after the July 2026 API price reduction

⚙️Input Parameters

messages

array

Chat or reasoning task messages sent to the OpenAI API or supported OpenAI product surface.

model

string

Use gpt-5.6-luna for the cost-sensitive GPT-5.6 tier in the OpenAI API.

💡Usage Examples

Example 1

Input Parameters

{
  "model": "gpt-5.6-luna",
  "messages": [
    {
      "role": "user",
      "content": "Classify whether this issue needs a cheap fix attempt or escalation to a stronger reasoning model."
    }
  ]
}

Output Results

A response from GPT-5.6 Luna. Escalate to Terra or Sol only when the same validation harness proves Luna is insufficient.

Quick Actions

Use NowView Documentation

Technical Specifications

Hardware Type: OpenAI-hosted
Commercial Use: Supported
Pricing: $0.20/1M input, $0.02/1M cached input, and $1.20/1M output tokens; prompts above 272K input use 2x input and 1.5x output rates; cache writes cost 1.25x uncached input
Platform: OpenAI

Related Keywords

GPT-5.6 Lunagpt-5.6-lunaOpenAI GPT-5.6GPT-5.6 Luna pricingGPT-5.6 Luna context windowlow-cost GPT-5.6GPT-5.6 API pricing

Related Models

GPT-5.6 Terra

OpenAI's generally available GPT-5.6 tier for workloads that balance intelligence and cost below the frontier Sol tier.

GPT-5.6 Sol

OpenAI's generally available frontier GPT-5.6 model for complex professional work, hard reasoning, coding, and scientific tasks.

Bielik 1.5B v3 Instruct

Bielik-1.5B-v3-Instruct is a generative text model featuring 1.6 billion parameters. It is result of collaboration between the open-science/open-souce project SpeakLeash and the High Performance Computing (HPC)