# OpenAI gates GPT-5.6 to 20 government-picked partners

> OpenAI previewed GPT-5.6 (Sol, Terra, Luna) to about 20 government-approved partners only, via API and Codex.

*OpenAI complied with a federal access gate in a day, then called it a mistake. Neat trick.*

By The InsidersFeed Desk · InsidersFeed
Canonical: https://insidersfeed.com/news/gpt-5-6-government-preview-gate

Here is the move. On 26 June OpenAI previewed its most capable model family yet — **GPT-5.6**, in three flavours called **Sol**, **Terra** and **Luna** — and then handed it to roughly **20 partners the US government picked**. Not you. Not most enterprises. Not ChatGPT users. Twenty names on a list nobody has published, reachable only through the **API and Codex**.

In the same breath, OpenAI said it **does not think this kind of government access process should become the long-term default**. Read that twice. The company complied with the gate on day one, shipped exactly as the rules demanded, and then publicly wished the rules away. That is not defiance. That is having it both ways — full compliance, plus a free marketing halo that says *our model is so dangerous the government made us lock it up*.

## Who are the 20? Nobody will say

The obvious question — **which 20 partners?** — has no answer. OpenAI says participation was **approved by the government**; it did not name the approvers or the approved. So the most powerful commercial AI in the world debuted as a private club with a secret guest list, blessed by an agency nobody has to identify. If a rival startup had done this, the same voices now nodding along would call it regulatory capture with extra steps.

> At the US government's request, OpenAI began with a limited preview to about 20 trusted partners whose participation was approved by the government, via the OpenAI API and Codex rather than ChatGPT, and expects to expand access in the coming weeks.
> — [VentureBeat](https://venturebeat.com/technology/openai-unveils-gpt-5-6-sol-terra-and-luna-models-but-only-accessible-to-limited-preview-partners-for-now-per-us-gov), 2026-06-26

To be fair to OpenAI, it did not invent the cage. The gate traces to a **Trump executive order of 2 June 2026** directing agencies to build a framework to benchmark and assess new models and designate **'covered frontier models'** — systems with advanced cyber capability — before wide release. This is the same federal machinery that yanked Anthropic's Fable 5 and Mythos 5 in June. The story here is that the gate is **no longer an Anthropic problem**. It is becoming the pattern across labs, and OpenAI just became the second big name to walk through it.

> OpenAI released its powerful new GPT-5.6 model under restrictions tied to a Trump executive order directing agencies to assess frontier models before wide deployment, starting with a limited set of government-approved partners.
> — [Axios](https://www.axios.com/2026/06/26/openai-gpt-sol-terra-luna-trump), 2026-06-26

## The rename is a price list in disguise

Ignore the astronomy branding for a second. The new naming scheme — **number = generation, names = durable capability tiers** — is genuinely tidy, but it is also a quiet repricing. **Terra** is pitched as roughly **GPT-5.5-level performance at about half the cost**. That is not a model launch; that is a discount wearing a new name. Sol is the halo product you can't have yet, Terra is the one they actually want your budget on, and Luna is the loss-leader that keeps you off a competitor's cheap tier.

Here is the ladder, per 1M tokens, input / output:

| Tier | Role | Price in / out | Preview status |
| --- | --- | --- | --- |
| **Sol** | Top flagship — hardest reasoning, coding, agentic, security | **$5 / $30** | Gov-gated, ~20 partners |
| **Terra** | Balanced everyday, ~GPT-5.5 quality at ~half cost | **$2.50 / $15** | Gov-gated |
| **Luna** | Budget — strong capability, lowest cost | **$1 / $6** | Gov-gated |

> **Key:** The tell: OpenAI shipped a full three-tier commercial price list for a family it says is too sensitive for general release. You do not publish per-token pricing for something you plan to keep locked up. Expansion 'in the coming weeks' was always the plan; the gate is the launch event.

## The cyber worry is real — credit where due

Now the part where OpenAI earns some respect. Under its **Preparedness Framework**, it classes Sol, Terra and Luna as **High capability in Cybersecurity and Biological/Chemical** — and says plainly that **none reach High in AI Self-Improvement**. On cyber specifically, OpenAI's read is that **Sol is better at helping people find and fix vulnerabilities than at reliably carrying out end-to-end attacks**, and that it does not cross the 'critical' threshold. That is a real, checkable safety disclosure, not vapour, and it is more than most labs volunteer.

So the fair verdict is split. The transparency is genuine and the cyber concern is not manufactured. But the *staging* — comply loudly, protest gently, publish the prices, expand in weeks — is a masterclass in turning a compliance obligation into a positioning win. OpenAI also trailed an **'ultra' mode** that splits work across sub-agents, and **Sol on Cerebras at up to 750 tokens/second in July** for select customers. Powerful, yes. Available to you, not yet.

> **Info:** Bottom line: GPT-5.6 is probably excellent and the safety work looks real. But 'the government made us gate it' is doing a lot of promotional heavy lifting for a company that complied without a fight and priced all three tiers on the way out the door.

## Key takeaways

- GPT-5.6 shipped 26 June 2026 in three tiers: Sol (flagship), Terra (~GPT-5.5 quality at half cost) and Luna (budget) — the number is the generation, the names are the tiers.
- The preview went to only ~20 partners the US government hand-picked, through the API and Codex, not ChatGPT — the first US flagship held behind a federal access gate.
- It follows a 2 June 2026 Trump executive order requiring agencies to benchmark 'covered frontier models' with advanced cyber capability before wide release.
- OpenAI complied instantly but insists the gate 'should not become the long-term default' — having it both ways: compliance plus a 'too powerful to ship' halo.
- Pricing (per 1M tokens): Sol $5/$30, Terra $2.50/$15, Luna $1/$6 — the rename doubles as a quiet price-tiering move.
- OpenAI rates all three High for cyber and bio/chem risk under its Preparedness Framework, but none reach High for AI self-improvement.

## FAQ

### Why is GPT-5.6 only available to some partners?
At the US government's request, OpenAI limited the 26 June preview to about 20 government-approved partners via the API and Codex, not ChatGPT. It stems from a 2 June 2026 executive order requiring agencies to benchmark 'covered frontier models' with advanced cyber capability before wide release. OpenAI says it expects to expand access in the coming weeks.

### What are Sol, Terra and Luna?
They are the three tiers of GPT-5.6. Sol is the top flagship for the hardest reasoning, coding, agentic and security tasks; Terra is a balanced everyday model at roughly GPT-5.5 quality for about half the cost; Luna is the budget tier. Under the new scheme the number is the generation and the names are durable capability tiers that advance on their own cadence.

### How much does GPT-5.6 cost?
Per 1M tokens (input / output): Sol is $5 / $30, Terra is $2.50 / $15, and Luna is $1 / $6. Terra is explicitly pitched as roughly GPT-5.5-level performance at about half the cost, which makes the rename double as a repricing.

### Does OpenAI support the government access gate?
No. OpenAI complied with the preview restriction but says it 'does not believe this kind of government access process should become the long-term default,' arguing it keeps the best tools from developers, enterprises, cyber defenders and global partners. Critics note the company complied instantly and still gained a 'too powerful to release' halo.

### Is GPT-5.6 a cybersecurity risk?
OpenAI rates all three tiers High capability in Cybersecurity and Biological/Chemical under its Preparedness Framework, but says none reach High in AI Self-Improvement. On cyber it says Sol is better at helping people find and fix vulnerabilities than at reliably running end-to-end attacks, and does not reach the 'critical' threshold.

### Is this the same rule that hit Anthropic?
Yes. The access gate traces to the same 2 June 2026 executive order and federal machinery that pulled Anthropic's Fable 5 and Mythos 5 in June. The significance is that it is no longer singling out Anthropic — the government-approved gate is becoming a pattern across major US labs.

## Sources

- [Previewing GPT-5.6 Sol: a next-generation model](https://openai.com/index/previewing-gpt-5-6-sol/) — OpenAI, 2026-06-26
- [OpenAI unveils GPT-5.6 Sol, Terra and Luna models — but only accessible to limited preview partners for now, per US Gov](https://venturebeat.com/technology/openai-unveils-gpt-5-6-sol-terra-and-luna-models-but-only-accessible-to-limited-preview-partners-for-now-per-us-gov) — VentureBeat, 2026-06-26
- [OpenAI releases powerful new GPT-5.6 model under restrictions](https://www.axios.com/2026/06/26/openai-gpt-sol-terra-luna-trump) — Axios, 2026-06-26