OpenAI announces GPT‑5.6 Sol, its next-generation flagship model beating Claude Mythos 5

OpenAI today announced a limited preview of its new GPT-5.6 model series, which includes the Sol, Terra, and Luna models targeting different price points.

GPT-5.6 Sol is the flagship model targeted at demanding reasoning and agentic workloads. GPT-5.6 Terra is positioned as a balanced model for everyday work, featuring performance competitive with GPT-5.5 while being half the cost. GPT-5.6 Luna is the fastest and most affordable model, delivering strong capability at a lower price point.

Unlike previous model releases from OpenAI, GPT-5.6 is starting with a limited preview for a small group of trusted partners due to U.S. government restrictions. As expected, OpenAI previewed its plans and the models' capabilities to the U.S. government ahead of launch, and the government asked OpenAI to limit the first wave of access to select partners. OpenAI also mentioned in the official announcement blog post that it does not believe this type of government access process should become the long-term default.

OpenAI highlighted that GPT-5.6 Sol comes with a robust safety stack featuring improved protections for higher-risk activity, sensitive cyber requests, and repeated misuse. The company also spent several weeks pressure-testing the system and hardening it against real-world attacks.

Sol is our new flagship and a step function better than GPT-5.5.

Terra delivers performance competitive to GPT-5.5 at 2x lower cost.

Luna is our most cost-efficient model, delivering strong capability at our lowest cost.

Together, the GPT-5.6 family gives people and developers… pic.twitter.com/Wm7wF3abHN

— OpenAI (@OpenAI) June 26, 2026

On the capability side, as expected, GPT-5.6 Sol is OpenAI’s strongest model yet. It delivers better results in agentic performance across coding, biology, and cybersecurity. On the Terminal-Bench 2.1 benchmark, which tests command-line workflows requiring planning, iteration, and tool coordination, GPT-5.6 Sol sets a new record with a score of 91.9%, beating Anthropic's Claude Mythos 5.

Additionally, GPT-5.6 introduces a new "max" reasoning effort for even deeper reasoning. The new "ultra" mode uses subagents to accelerate complex work beyond what a single agent can handle.

Pricing starts at $5 per million input tokens and $30 per million output tokens for Sol. Terra costs $2.50 for input and $15 for output, while Luna costs $1 for input and $6 for output. GPT-5.6 comes with more predictable prompt caching, including support for explicit cache breakpoints and a 30-minute minimum cache life. Sol will also launch on Cerebras in July at speeds up to 750 tokens per second for select customers.

OpenAI plans to make GPT-5.6 Sol, Terra, and Luna broadly available in ChatGPT, Codex, and the API in the coming weeks.