Introducing GPT‑5.3‑Codex
Expanding Codex across the full spectrum of professional work on a computer.
We’re introducing a new model that unlocks even more of what Codex can do: GPT‑5.3‑Codex, the most capable agentic coding model to date. The model advances both the frontier coding performance of GPT‑5.2‑Codex and the reasoning and professional knowledge capabilities of GPT‑5.2, together in one model, which is also 25% faster. This enables it to take on long-running tasks that involve research, tool use, and complex execution. Much like a colleague, you can steer and interact with GPT‑5.3‑Codex while it’s working, without losing context.
GPT‑5.3‑Codex is our first model that was instrumental in creating itself. The Codex team used early versions to debug its own training, manage its own deployment, and diagnose test results and evaluations—our team was blown away by how much Codex was able to accelerate its own development.
With GPT‑5.3‑Codex, Codex goes from an agent that can write and review code to an agent that can do nearly anything developers and professionals can do on a computer.
Frontier agentic capabilities
GPT‑5.3‑Codex sets a new industry high on SWE-Bench Pro and Terminal-Bench, and shows strong performance on OSWorld and GDPval, four benchmarks we use to measure coding, agentic and real-world capabilities.
Coding
GPT‑5.3‑Codex achieves state-of-the-art performance on SWE-Bench Pro, a rigorous evaluation of real-world software engineering. Where SWE‑bench Verified only tests Python, SWE‑Bench Pro spans four languages and is more contamination‑resistant, challenging, diverse and industry-relevant. It also far exceeds the previous state-of-the-art performance on Terminal-Bench 2.0, which measures the terminal skills a coding agent like Codex needs. Notably, GPT‑5.3‑Codex does so with fewer tokens than any prior model, letting users build more.
| SWE-Bench Pro (Public) | Accuracy |
|---|---|
| GPT-5.2 | 30% |
| GPT-5.2-Codex | 40% |
| GPT-5.3-Codex | 56.8% |
| Terminal-Bench 2.0 | Accuracy |
|---|---|
| GPT-5.3-Codex | 77.3% |
| GPT-5.2-Codex | 64.0% |
| GPT-5.2 | 62.2% |
Web development
Combining frontier coding capabilities, improvements in aesthetics, and compaction results in a model that can do striking work, building highly functional complex games and apps from scratch over the course of days. To test the model’s web development and long-running agentic capabilities, we asked GPT‑5.3‑Codex to build us two games: version two of the racing game from the Codex app launch, and a diving game. Using the develop web game skill and preselected, generic follow-up prompts like "fix the bug" or "improve the game", GPT‑5.3‑Codex iterated on the games autonomously over millions of tokens.
Racing Game
A racing game, complete with different racers, eight maps, and even items to use with the space bar. Play it for yourself here!
Diving Game
A diving game where you explore various reefs, collect them all to complete your fish codex, all the while managing oxygen, pressure, and hazards. Play it for yourself here!
GPT‑5.3‑Codex also better understands your intent when you ask it to make day-to-day websites, compared to GPT‑5.2‑Codex. Simple or underspecified prompts now default to sites with more functionality and sensible defaults, giving you a stronger starting canvas to bring your ideas to life.
Beyond coding
Software engineers, designers, product managers, and data scientists do far more than generate code. GPT‑5.3‑Codex is built to support all of the work in the software lifecycle—debugging, deploying, monitoring, writing PRDs, editing copy, user research, tests, metrics, and more. Its agentic capabilities go beyond software, helping you build whatever you want to build—whether it’s slide decks or analyzing data in sheets.
With custom skills similar to those used for our previous GDPval results, GPT‑5.3‑Codex also shows strong performance on professional knowledge work as measured by GDPval, matching GPT‑5.2. GDPval is an evaluation OpenAI released in 2025 that measures a model’s performance on well‑specified knowledge‑work tasks across 44 occupations.
Sample task
Prompt + task context
You are a financial advisor working at a wealth management firm. You have been tasked to create a 10-slide PowerPoint presentation to share talking points on why financial advisors, as fiduciaries, should strongly recommend against certain investment decisions.
GPT-5.3-Codex output
GPT‑5.3‑Codex demonstrates far stronger computer use capabilities than previous GPT models.
Securing the cyber frontier
Over recent months, we’ve seen meaningful gains in model performance on cybersecurity tasks, benefiting both developers and security professionals. GPT‑5.3‑Codex is the first model we classify as High capability for cybersecurity-related tasks under our Preparedness Framework.
Availability & details
GPT‑5.3‑Codex is available with paid ChatGPT plans, everywhere you can use Codex: the app, CLI, IDE extension and web. We are working to safely enable API access soon.
With this update, we are also now running GPT‑5.3‑Codex 25% faster for Codex users, thanks to improvements in our infrastructure and inference stack, resulting in faster interactions and faster results.
What’s next
With GPT‑5.3‑Codex, Codex is moving beyond writing code to using it as a tool to operate a computer and complete work end to end. By pushing the frontier of what a coding agent can do, we’re also unlocking a broader class of knowledge work—from building and deploying software to researching, analyzing, and executing complex tasks.