OpenAI Launches GPT-6 Sol and Luna with 50% API Price Cuts

OpenAI launched GPT-6 Sol and GPT-6 Luna on September 22, 2026, expanding its GPT-6 model family with two lower-cost API options that cut token prices by 50% compared to predecessor models. Positioned directly below the flagship GPT-6 Astra model that debuted earlier in the month, Sol and Luna bring frontier capabilities to daily developer workflows at significantly reduced operating costs. Both models are immediately available to developers through OpenAI's API under the model identifiers gpt-6-sol and gpt-6-luna.

The release targets software developers and enterprise teams who require scalable AI reasoning without paying top-tier pricing for routine computational tasks. OpenAI trained both GPT-6 Sol and GPT-6 Luna using optimization techniques similar to those applied to GPT-6 Astra. The result is a pair of API-only models designed to balance intelligence, processing latency, and execution cost across diverse software workflows.

API Integration and Code Implementation

Developers can begin calling the new models immediately using standard OpenAI SDK pipelines. Because GPT-6 Sol and GPT-6 Luna are API-only releases, integration requires updating the target model parameter in existing API request handlers. Below is a complete Python example using the official OpenAI client library to send a completion request to the gpt-6-sol endpoint:

from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="gpt-6-sol",
    messages=[
        {"role": "system", "content": "You are an expert software engineer."},
        {"role": "user", "content": "Write a Python function to parse JSON safely."}
    ]
)

print(response.choices[0].message.content)

To route lower-complexity tasks to the faster, budget-friendly Luna model, developers simply swap the model parameter string to gpt-6-luna in their request configuration:

from openai import OpenAI

client = OpenAI()

response = client.chat.completions.create(
    model="gpt-6-luna",
    messages=[
        {"role": "system", "content": "You are a lightweight developer assistant."},
        {"role": "user", "content": "Summarize these system status logs."}
    ]
)

print(response.choices[0].message.content)

Meet GPT-6 Sol and Luna, two models that bring frontier intelligence to everyday work with different balances of capability and cost.

OpenAI

API Pricing Reductions and Hardware Execution Speeds

Related: OpenAI Launches GPT-6 Astra, Sol, and Luna with 50% Price Cuts

The primary highlight of the GPT-6 Sol and Luna release is API cost reduction. OpenAI confirmed that both new models feature token price cuts of 50% or more when compared against earlier promotional rates for the GPT-5.6 generation. Primary OpenAI documentation for GPT-5.6 Sol listed pricing at $5.00 per 1 million input tokens and $30.00 per 1 million output tokens, while GPT-5.6 Luna input pricing was listed at $1.00 per 1 million tokens.

However, secondary sources state differing baseline numbers for the prior model generation. According to alternative reports, GPT-5.6 promotional pricing stood at $4.00 input and $20.00 output per 1 million tokens for Sol, alongside $0.20 input and $1.20 output per 1 million tokens for Luna. Despite these conflicting reports regarding baseline numbers, the new release effectively halves the token expense for developers transitioning to GPT-6 Sol and Luna.

In terms of hardware execution speed, OpenAI previously demonstrated fast throughput capabilities with earlier model variations. During hardware testing in July, GPT-5.6 Sol achieved processing speeds up to 750 tokens per second when executing on specialized Cerebras hardware infrastructure. While OpenAI has not published hardware-specific execution benchmarks for the newly released GPT-6 Sol, the underlying architectural improvements aim to maintain fast inference across enterprise workloads.