OpenAI released GPT-5.5 on April 23rd to ChatGPT Plus, Pro, Business, and Enterprise subscribers, marking the company's fastest turnaround between major model releases to date. The launch came just six weeks after GPT-5.4 and signals how aggressively OpenAI is iterating to hold off competitors closing in from multiple directions.
OpenAI president Greg Brockman described the new model as a meaningful step toward more agentic and intuitive computing, framing it as part of a longer journey toward a unified AI "super app." Unlike previous releases focused narrowly on chat performance, GPT-5.5 is built to handle extended, multi-step tasks without constant human hand-holding.
The Upgrade That Actually Matters for Business
The gains in GPT-5.5 are concentrated where business users actually feel friction. The model is stronger at writing and debugging code, operating software, researching across multiple sources, analyzing data, and producing formatted outputs like documents and spreadsheets. Most importantly, it can work through ambiguous, multi-part tasks by planning its own steps, catching errors, and continuing until the job is done.
This agentic capability has been the missing piece for enterprise adoption. Previous models were excellent at answering questions or generating first drafts. GPT-5.5 is designed to finish the whole task. OpenAI reported that more than 85% of its own employees now use Codex every week across departments including finance, communications, marketing, and data science.
On benchmark tests, GPT-5.5 posted an 82.7% score on Terminal-Bench 2.0, outperforming Claude Opus 4.7 and Gemini 3.1 Pro across most categories. It also showed notable gains on scientific research evaluations, with teams already using it to analyze biochemical datasets and stress-test complex arguments across multi-pass research workflows.
The Bigger Context
The release is not just a product update. OpenAI now reports 900 million weekly active users and more than 50 million paying subscribers. The company is under real pressure from Anthropic, which has been gaining ground in enterprise, and from the open-source threat represented by DeepSeek's new V4 model released the same week.
API access to GPT-5.5 was withheld at launch and opened to developers the following day, with OpenAI noting that serving it at scale required different safety infrastructure than consumer deployment. The company is running a Trusted Access for Cyber program, giving verified security researchers broader access to advanced cybersecurity features.
GPT-5.5 is available at standard ChatGPT pricing for paid subscribers, with GPT-5.5 Pro included at the Pro tier. Codex users also have access through the AI agent platform. OpenAI positions this release as one step in a longer sequence leading toward systems that can operate computers on behalf of users with precision and reliability.
For the executives I speak with on LinkedIn, the question is no longer whether to use AI tools in daily workflows. It is which tools can actually execute a full task versus which ones still require a babysitter to get results. GPT-5.5 is a real answer to that question.
