GPT-5.6: Code Faster, Secure Better, Act Now
OpenAI's GPT-5.6 Sol is the most powerful model for coding, cybersecurity, and agentic tasks.
Jul 10, 2026 (Updated Jul 10, 2026) - Written by Christian Tico
This image is part of OpenAI's official brand assets, available from their press kit
Why Your Brilliant Ideas Keep Failing (And How to Fix It)
Stop launching marketing campaigns or video concepts based on pure guesswork. Run your concepts through the AI Idea Evaluation Panel to get instant reports on strengths and weaknesses.
OpenAI has officially launched GPT-5.6, marking a monumental leap in artificial intelligence capabilities with its strongest model ever designed for coding, cybersecurity, and agentic workflows. This new family of models, comprising the flagship Sol, the balanced Terra, and the budget-friendly Luna, is now rolling out globally across ChatGPT, Codex, and the OpenAI API. The rollout comes after a brief 12-day delay triggered by U.S. government reviews focused on national security concerns regarding the potential misuse of advanced AI technologies.
The GPT-5.6 Family: Sol, Terra, and Luna
GPT-5.6 introduces three distinct capability tiers designed to meet diverse user needs, from enterprise-grade operations to everyday tasks. Each model is optimized for specific use cases while maintaining the core advancements of the generation.
- Sol: The flagship model engineered for complex agentic tasks and high-performance demands. It is 54% more token-efficient than previous versions when handling AI coding tasks.
- Terra: An intermediate option that delivers competitive performance with GPT-5.5 at a lower cost, making it ideal for everyday production work.
- Luna: The fastest and most affordable model in the family, perfect for budget-conscious users needing rapid responses.
CEO Sam Altman has emphasized that these models are orders of magnitude more efficient and cost-effective than their predecessors, setting a new standard for performance in the AI industry.
Revolutionizing Cybersecurity and Coding
GPT-5.6 Sol stands out as the strongest cybersecurity model OpenAI has ever created, achieving frontier performance with significantly fewer tokens. This advancement enables the model to support critical defensive activities including threat modeling, code review, patching, and blue teaming simulations. In blue teaming, the model simulates attacks on a user's own systems to identify weaknesses before real hackers can exploit them.
In the realm of coding, the model shifts the performance-efficiency frontier for long-horizon security tasks such as vulnerability research and exploitation. Benchmark evaluations on ExploitGym, created by UC Berkeley researchers in collaboration with OpenAI, demonstrate strong improvements in cyber capabilities as reasoning depth increases. The model's capabilities extend beyond coding to biology, offering powerful tools for scientific research and enterprise applications.
Global Rollout and New Agent: ChatGPT Work
The rollout of GPT-5.6 is starting globally on Thursday, July 9, 2026, and will continue gradually toward full availability over the next 24 hours. Access may be staged by account tier, with immediate availability for Plus, Pro, Business, and Enterprise users. Free ChatGPT users on Mac and Windows will also have immediate access via the desktop app, while mobile and web users with Pro, Enterprise, and Edu plans will receive access first, followed by Plus and Business users within the next few days.
Introducing ChatGPT Work
To celebrate the launch, OpenAI unveiled a new AI agent called ChatGPT Work. This agent combines the capabilities of ChatGPT and Codex, allowing non-technical users to leverage Codex's power for non-coding tasks. It is powered by the GPT-5.6 model suite and is accessible to millions of users worldwide.
Pricing and Technical Enhancements
GPT-5.6 is priced per 1 million tokens, offering clear cost structures for developers and enterprises. Sol is priced at $5 for input and $30 for output, Terra at $2.50 for input and $15 for output, and Luna at $1 for input and $6 for output. The new generation also introduces more predictable prompt caching, including support for explicit cache breakpoints and a 30-minute minimum cache life. Cache writes are billed at 1.25 times the uncached input rate, while cache reads receive a 90% discount. Additionally, GPT-5.6 Sol will run on Cerebras hardware at speeds up to 750 tokens per second in July, bringing frontier intelligence to customers at unprecedented speeds.
Conclusion: A New Era of AI Intelligence
With the public release of GPT-5.6, OpenAI has delivered a transformative tool that redefines the boundaries of artificial intelligence. Its unparalleled capabilities in coding, cybersecurity, and agentic tasks position it as the most powerful model in the company's history. As the rollout continues globally, users from all sectors will gain access to this cutting-edge technology, unlocking new possibilities for innovation and security in the digital age.
The true breakthrough of GPT-5.6 is not its offensive prowess, which remains intentionally capped at finding vulnerabilities rather than executing full-chain attacks, but its unprecedented efficiency in defensive security, proving that the most powerful AI tool for cybersecurity is actually a force multiplier for protection rather than destruction. This shift redefines the industry's safety paradigm by demonstrating that frontier intelligence can be engineered to close the gap between vulnerability discovery and remediation faster than the gap between exploitation and defense.
How much does GPT-5.6 cost per million tokens?