OpenAI’s GPT-5.6 family becomes generally available on Thursday, July 9, 2026, introducing three tiers, Sol, Terra, and Luna, across ChatGPT, the API, and Codex. The release follows a late-June preview limited to a small group of trusted partners coordinated with the U.S. government.
Here is what is arriving on day one.
The GPT-5.6 Family: Sol, Terra, and Luna
GPT-5.6 introduces a new naming system. The number identifies the generation. Sol, Terra, and Luna are durable capability tiers that can advance on their own cadence over time.
- Sol is the flagship, built for deep reasoning and agentic work.
- Terra is positioned as a balanced everyday model, delivering competitive performance with GPT-5.5 at roughly half the price.
- Luna is the fast and most affordable tier, designed for high-volume workloads where cost and latency matter most.
What New Reasoning Controls Does GPT-5.6 Ship With?
OpenAI is shipping two new modes alongside the family. A new max reasoning effort gives Sol the most time to reason deeply before answering. A new ultra mode goes beyond a single agent by spinning up subagents to parallelize complex tasks and accelerate completion. Together, they let developers dial up depth or breadth depending on the job.
Where Does GPT-5.6 Lead?
OpenAI is anchoring its launch around three capability areas where the benchmarks moved sharply: coding, biology, and cybersecurity.
Coding
GPT-5.6 Sol records the highest score reported on Terminal-Bench 2.1, a benchmark built around command-line workflows that require planning, iteration, and tool coordination. For developer teams running agents in real terminal environments, this is the headline result.
Biology
On GeneBench v1, which evaluates long-horizon genomics and quantitative-biology analyses, GPT-5.6 Sol achieves stronger results than GPT-5.5 while using fewer tokens. That combination of higher quality and lower token spend is exactly what research teams look for.
Cybersecurity
GPT-5.6 Sol is the most capable cybersecurity model OpenAI has shipped. On ExploitBench it is competitive with Mythos Preview while using only about one-third of the output tokens. On ExploitGym, a benchmark created by UC Berkeley researchers in collaboration with OpenAI and other frontier labs, all three tiers show strong gains as reasoning effort increases.
The Strongest Safety Stack So Far
OpenAI describes GPT-5.6’s safeguards as the most robust it has released, with configurations matched to each tier’s capabilities. The stack includes:
- Refusal training tuned for prohibited cyber assistance, including attempts to disguise intent or jailbreak the model.
- Real-time cyber and biology misuse classifiers that evaluate output as it is generated. For higher-risk cases, generation can be paused while a larger reasoning model reviews the conversation.
- Account-level review that can be triggered by flagged activity.
- Differentiated access matched to model capability.
- Over 700,000 A100-equivalent GPU hours of automated red-teaming aimed at finding universal jailbreaks, plus third-party human red-teaming continuing through the preview period and a rapid-response process for newly discovered attacks.
OpenAI also notes that GPT-5.6 Sol does not cross the Cyber Critical threshold under its Preparedness Framework. In evaluations involving Chromium and Firefox, it identified bugs and exploitation primitives but did not autonomously produce a functional full-chain exploit under the conditions tested. OpenAI frames the model as better at helping people find and fix vulnerabilities than at reliably carrying out end-to-end attacks.
The Preview Process and Government Coordination
The late-June preview was coordinated with the U.S. government. At the government’s request, OpenAI started with a small group of trusted partners whose participation was shared with the Administration, before moving to broader availability. OpenAI has been clear that it does not want this kind of government access process to become the long-term default, arguing that it keeps the best tools away from users, developers, enterprises, cyber defenders, and global partners who need them. The company describes the current arrangement as a short-term step taken while it works with the Administration to develop a cyber Executive Order framework and a repeatable process for future model releases.
How Is GPT-5.6 Priced?
GPT-5.6 is priced per million tokens across the three tiers:
- Sol: $5 input / $30 output
- Terra: $2.50 input / $15 output
- Luna: $1 input / $6 output
The release also introduces more predictable prompt caching. Developers get explicit cache breakpoints, a 30-minute minimum cache life, and clearer billing: cache writes are charged at 1.25x the uncached input rate, while cache reads continue to receive the 90% cached-input discount. For teams running long sessions or agentic loops, that makes cost planning far more straightforward.
What Does Day-One Availability Look Like?
With general availability opening tomorrow, GPT-5.6 Sol, Terra, and Luna roll out through the API and Codex to a wider audience, alongside broader ChatGPT access. OpenAI is also launching GPT-5.6 Sol on Cerebras at up to 750 tokens per second during July, with access initially limited to select customers. That hardware path puts the flagship model in a different latency bracket for teams willing to route specific workloads through it.
For developers and businesses, tomorrow is the day to benchmark the three tiers against your real workloads. GPT-5.6 Sol is built to shine at coding, long-horizon biology, and security-heavy agentic loops, Terra offers the strongest price-to-performance ratio for everyday tasks, and Luna enables high-volume use cases that were previously uneconomical.
FAQ
When does GPT-5.6 launch?
GPT-5.6 reaches general availability on Thursday, July 9, 2026, following a late-June preview limited to a small group of trusted partners.
What are the Sol, Terra, and Luna tiers?
Sol is the flagship built for the deepest reasoning and toughest agentic work, Terra is a balanced everyday model that matches GPT-5.5 performance at roughly half the price, and Luna is a fast, most affordable tier designed for high-volume workloads.
How much does GPT-5.6 cost?
Pricing is per million tokens: Sol at $5 input / $30 output, Terra at $2.50 input / $15 output, and Luna at $1 input / $6 output, with cache writes billed at 1.25x the uncached input rate and cache reads receiving a 90% discount.
Want help with this in Myrtle Beach or Long Island? Talk to us.
NewSunSEO
Grow in Google and AI Search
AI-SEO, local search, paid ads and websites from one team, with offices on Long Island and in Myrtle Beach.
- AI-SEO
- Local AI-SEO
- AI Consulting
- PPC
- Web Design
- Web Development
- Long Island Office(631) 320-1700
- Myrtle Beach Office(843) 492-7131
Request a Free Proposal
Tell us about your business and where you want to grow.
We will reply by email to talk through your goals and what a proposal would include.
We use your details only to reply to you and send what you asked for. See our Privacy Policy.
