GPT-5.6 for Business: Tiers, Pricing, and Access

What small and midsize businesses should know before they plan for OpenAI's GPT-5.6 tiers.

Restricted as of July 1, 2026: GPT-5.6 (Sol, Terra, and Luna) is in a limited preview open only to a small set of vetted organizations, via the API and Codex, at the US government's request; general availability is planned in the coming weeks (OpenAI). GPT-5.6 for business is OpenAI's newest model family, split into three tiers. This guide explains each tier in plain terms for small and midsize teams.

You cannot buy general access to GPT-5.6 today. But you can plan now so you are ready when it opens. We map the tiers, the prices, and the compliance steps to real business tasks.

This page focuses on facts you can act on. It avoids hype and sticks to what OpenAI has published. Use it to decide where GPT-5.6 might fit once access widens.

Reviewed by Jonathan West, Founder of Layer3 Labs, on July 1, 2026. We research using primary vendor and regulator sources.

What GPT-5.6 for business means today

GPT-5.6 for business is a three-tier model family that OpenAI announced on June 26, 2026. The tiers are named Sol, Terra, and Luna. Each tier targets a different mix of task difficulty and cost.

The family is not open to everyone yet. Access is a limited preview for a small set of vetted organizations. It runs through the API and Codex only, at the US government's request.

For most SMBs, this means planning rather than buying. You can study the tiers now. Then you can request access when general availability arrives in the coming weeks.

  • Sol: hardest problems, including complex coding and security research.
  • Terra: high-volume business tasks like support, internal tools, and document analysis.
  • Luna: fast, low-cost everyday work like summarizing, drafting, and routine automation.
Sources
OpenAI — Previewing GPT-5.6 Sol

Deciding how GPT-5.6 tiers like Sol, Terra, and Luna fit your business without overspending or risking regulated data? Layer3 Labs can map them to your workflows and confirm the compliance steps before you commit.

Book a Consultation

Match GPT-5.6 tiers to your business tasks

Pick the GPT-5.6 tier that fits the job, not the most powerful one. Most business work does not need the top tier. Matching the tier to the task keeps quality high and cost low.

Sol is built for the hardest problems. It adds new "max" and "ultra" reasoning modes. The "ultra" mode uses subagents to speed up complex work.

Terra suits steady, high-volume work. Think customer support, internal tools, and document analysis. Luna is the fast, cheap option for summaries, drafts, and routine automation.

  • Use Luna first for simple drafting and summaries; it is the cheapest tier.
  • Move to Terra when volume grows or documents get longer.
  • Reserve Sol for deep coding or research where cheaper tiers fail.
A common SMB mistake is running every task on the top tier. Routing routine drafts to Luna and reserving Sol for hard problems can cut a monthly bill by a large share, since Sol output costs five times Luna output per token.

GPT-5.6 for business pricing by tier

GPT-5.6 pricing is set per million tokens and differs by tier. Tokens are small chunks of text the model reads and writes. Input tokens are what you send; output tokens are what the model returns.

Sol costs 5 dollars per million input tokens and 30 dollars per million output tokens. Terra costs 2.50 dollars input and 15 dollars output. Luna costs 1 dollar input and 6 dollars output.

Output tokens cost more than input tokens on every tier. So long, wordy answers drive most of your bill. Keeping prompts and answers tight lowers cost fast.

  • Sol: 5 dollars input / 30 dollars output per million tokens.
  • Terra: 2.50 dollars input / 15 dollars output per million tokens.
  • Luna: 1 dollar input / 6 dollars output per million tokens.
Sources
OpenAI — Previewing GPT-5.6 Sol

How to get GPT-5.6 access

GPT-5.6 access is currently a limited preview, not a public launch. OpenAI opened it to a small set of vetted organizations only. Access runs through the API and Codex at the US government's request.

General availability is planned in the coming weeks. Until then, you cannot self-serve the model on a standard plan. Plan your use cases so you can move quickly once it opens.

A useful step now is to keep a stable model you can already deploy. That way your project does not stall while you wait for preview access to widen.

  • Preview is limited to vetted organizations via API and Codex.
  • General availability is planned in the coming weeks (OpenAI).
  • Draft your use cases and data rules now, before access opens.
Sources
9to5Mac — OpenAI GPT-5.6 limited release

How to prepare your business now

You can prepare for GPT-5.6 before you have access. Start by listing the tasks you want to automate. Then sort each task by difficulty so you can map it to a tier later.

Write down which data each task touches. Mark anything regulated, like health or financial records. This makes the compliance step faster when access opens.

Run a small pilot on a model you can use today. A pilot proves the workflow and the value. Then you can swap in GPT-5.6 when it becomes available.

  • List candidate tasks and rank them easy, medium, or hard.
  • Map easy tasks to Luna, medium to Terra, and hard to Sol.
  • Flag which tasks involve regulated or sensitive data.
  • Pilot the workflow now on an available model to prove value.

GPT-5.6 for business compliance and data rules

GPT-5.6 inherits OpenAI's platform compliance coverage once a model is in scope. That coverage includes SOC 2, ISO 27001, and a HIPAA Business Associate Agreement on the API and Enterprise plans. These controls support many business and regulated needs.

The key caveat is that GPT-5.6 is still a preview. A model in preview may not yet be named in your signed BAA. Confirm the specific model is listed in your BAA before you send any regulated data.

If the model is not yet in your BAA, keep regulated data off it for now. Use it for non-sensitive tasks until the paperwork catches up. This one check prevents a costly compliance gap.

  • Coverage includes SOC 2, ISO 27001, and a HIPAA BAA on API and Enterprise.
  • Preview models may not be named in your BAA yet.
  • Confirm the exact model in your BAA before sending regulated data.
Confirm-BAA-in-preview: platform certifications do not automatically cover a new preview model. A signed BAA lists the specific models in scope, so a preview tier can sit outside it until OpenAI adds it. Verify the model name in writing first.
Sources
OpenAI — Previewing GPT-5.6 Sol

Recent Developments: GPT-5.6 Capabilities and Model Preview

GPT-5.6 capabilities now include enhanced cyber safeguards and a layered security stack, as announced in OpenAI's June 2026 preview of the Sol, Terra, and Luna models. OpenAI highlighted improved robustness from automated red-teaming and focused its limited preview on new security measures designed to address enterprise and business needs.

Each model within the GPT-5.6 series serves distinct priorities: Sol is the flagship for advanced tasks, Terra balances cost and performance at roughly half the price of GPT-5.5, and Luna offers the lowest-cost option while maintaining solid AI features. These updates allow businesses to select models aligned with their budget and risk requirements, with OpenAI emphasizing increased safety due to enhanced testing and feature gating through multiple safeguard layers (source: OpenAI).

  • OpenAI released a limited preview of GPT-5.6 featuring Sol, Terra, and Luna models.
  • Security is central, with a layered safeguard stack and automated red-teaming for robustness.
  • Sol is the flagship; Terra and Luna provide more affordable options with balancing features.
  • Businesses can match model selection to cost, performance, and risk tolerance.
  • The preview adds specialized cyber capabilities for enterprise and business deployments.
The latest GPT-5.6 preview prioritizes layered safeguards and cyber resilience, which OpenAI stresses are critical for enterprise users evaluating new AI model rollouts.
Sources
Primary sourceMatt Wolfe: AI News: Fable's Back But This New Model is Better?

Frequently Asked Questions

  • No. GPT-5.6 is in a limited preview open only to a small set of vetted organizations. It runs through the API and Codex at the US government's request. General availability is planned in the coming weeks.
  • The three tiers are Sol, Terra, and Luna. Sol handles the hardest problems like complex coding and security research. Terra fits high-volume business tasks, and Luna is the fast, low-cost tier for everyday work.
  • Pricing is per million tokens and varies by tier. Sol costs 5 dollars input and 30 dollars output. Terra costs 2.50 dollars input and 15 dollars output. Luna costs 1 dollar input and 6 dollars output.
  • Most SMBs should start with Luna, the cheapest tier, for drafting and summaries. Move to Terra for high-volume support and document work. Reserve Sol for deep coding or research where cheaper tiers fall short.
  • It can be, but you must verify first. GPT-5.6 inherits OpenAI's HIPAA BAA coverage on the API and Enterprise plans once a model is in scope. Because it is still a preview, confirm the specific model is named in your BAA before sending regulated data.
  • The ultra mode speeds up complex work by using subagents. It is part of the Sol tier, which also adds a max mode. These modes target the hardest coding and research problems.
  • Prepare now so you can move fast later. List your tasks, rank them by difficulty, and map each to a tier. Flag regulated data and run a pilot on a model you can already use to prove value.
  • Check OpenAI's own preview announcement for tiers and pricing. The OpenAI page for GPT-5.6 Sol is the primary source. This guide links to it so you can confirm every claim.

Plan your GPT-5.6 rollout with Layer3 Labs

Book a free 30-minute AI workflow audit with Layer3 Labs. We map GPT-5.6 tiers, cost, and compliance to your real tasks so you are ready when access opens.

Book Your Free Audit