Realtime AI News
OpenAI Previews GPT-5.6 Sol: Next-Generation Model with Stronger Coding, Science, and Security Capabilities
OpenAI previewed GPT-5.6 Sol on June 26, describing it as a next-generation model with enhanced capabilities in coding, science, and cybersecurity, paired with its most advanced safety stack.
On June 26, 2026, OpenAI published a preview of its next-generation model, GPT-5.6 Sol, via its official blog. The model is positioned as a frontier AI with significantly improved capabilities across three key domains: coding, scientific research, and cybersecurity.
OpenAI stated that GPT-5.6 Sol is equipped with its most advanced safety stack to date, designed to mitigate risks associated with cutting-edge AI systems. This continues OpenAI's tradition of releasing safety evaluations alongside frontier model announcements.
As of the preview, OpenAI has not disclosed a specific release date or pricing for GPT-5.6 Sol. The announcement arrives amid intensifying competition in the AI industry, with rivals such as Anthropic and Google DeepMind also accelerating their frontier model efforts.
The preview comes directly from OpenAI's official newsroom (openai.com), making it a first-party source with high credibility. Notably, within hours of the preview, multiple media outlets reported that the U.S. government had requested OpenAI to delay the model's public deployment, suggesting that GPT-5.6 Sol's safety implications have already become a policy-level concern.
Why it matters
GPT-5.6 Sol represents OpenAI's latest leap in coding, science, and cybersecurity capabilities, but government scrutiny over its deployment may set a precedent for how frontier AI models are released.