OpenAI expands GPT-5.6 availability and launches ChatGPT Work for enterprise productivity

OpenAI has transitioned its GPT-5.6 model suite to a public release following regulatory approval. The company also launched ChatGPT Work, an AI agent designed for enterprise integration.

AI & Culture Desk
Updated Nov 13, 2026 2 min read
OpenAI expands GPT-5.6 availability and launches ChatGPT Work for enterprise productivity

OpenAI has moved its GPT-5.6 model suite to a public release following a period of government-restricted access. The rollout comes after the Trump administration cleared the models, which were previously limited to government-approved organizations. CEO Sam Altman described the release as the company's best model to date. The suite includes three distinct variants: Sol, characterized as the workhorse, Terra, an intermediate option, and Luna, the budget-friendly tier. According to TechCrunch, the company is positioning these models to compete directly with rivals like Anthropic, specifically targeting enterprise workflows, coding, and scientific research.

Alongside the model update, OpenAI introduced ChatGPT Work, an AI agent designed to function as a workplace companion. According to The Verge, the tool combines ChatGPT with Codex, enabling non-technical users to utilize coding-like capabilities for tasks such as drafting documents, spreadsheets, and presentations. The agent features a unified plugins directory that connects to common business software, including Slack, Gmail, Google Drive, and various CRM platforms. OpenAI claims the tool can gather context from these files and workflows to generate finished materials, positioning it as a direct competitor to Anthropic’s Claude Cowork.

OpenAI is heavily marketing the cybersecurity capabilities of the GPT-5.6 family, which the company claims is its strongest cybersecurity model yet. TechCrunch reports that the models support defensive operations like threat modeling, code review, and blue teaming. These capabilities were a primary focus of the regulatory scrutiny that previously delayed the public release, with officials raising concerns over potential misuse. Despite these concerns, OpenAI asserts that the models achieve frontier performance while requiring fewer tokens, which the company frames as a necessary efficiency gain in an industry facing rising operational costs.

To justify the release, OpenAI is leaning on the Artificial Analysis Coding Agent Index to claim superiority over Anthropic. The company asserts that Sol, its most powerful model, sets a new state of the art score of 80, which it claims is 2.8 points higher than Anthropic’s Fable 5. OpenAI further argues that Sol achieves this while using less than half the output tokens and costing about one-third less. According to TechCrunch, Altman has stated that Sol is 54% more token-efficient for coding tasks, a metric the company hopes will appeal to cost-conscious enterprise customers.

Global availability for the new tools varies by user tier. The Verge reports that Mac and Windows users on the ChatGPT desktop app have immediate access to GPT-5.6 and ChatGPT Work. For web and mobile users, access is being prioritized for Pro, Enterprise, and Edu accounts, with Plus and Business users expected to receive access over the coming days. Pricing for the GPT-5.6 API is tiered by model: Sol is priced at $5 per million input tokens and $30 per million output tokens; Terra at $2.50 input and $15 output; and Luna at $1 input and $6 output.

Sources

  1. 01 OpenAI rolls out GPT-5.6 after government greenlight — and announces ‘ChatGPT Work’ — The Verge
  2. 02 OpenAI launches its new family of models with GPT-5.6 — TechCrunch