GPT-5.3-Codex
OpenAI's GPT-5.3-Codex is an advanced agentic coding model designed for complex software engineering tasks. It integrates professional knowledge with reasoning capabilities and supports input from text, images, and files.
GPT-5.3-Codex specs and pricing
- Provider: OpenAI
- Released: February 24, 2026
- Context window: 400K tokens
- Max output: 128K tokens
- Input: text, images, files
- Output: text
- Price: $1.75 input / $14 output per 1M tokens
- Reasoning: yes
- Tool calling: yes
What GPT-5.3-Codex is good at
- Complex software engineering and development projects
- Analyzing large codebases and technical documentation
- Automated reasoning tasks requiring multi-modal input
To-dos to try with GPT-5.3-Codex
- Analyze this 300-page technical specification document and generate a modular architecture plan.
- Review these screenshots of a UI design and write the corresponding frontend code components.
- Debug this repository of files and suggest a refactoring strategy to improve performance.
FAQ
- What is the primary purpose of GPT-5.3-Codex? It is an agentic coding model developed by OpenAI specifically for software engineering and complex reasoning tasks.
- Can I upload images to this model? Yes, the model supports both image and file inputs alongside standard text.
- How much does GPT-5.3-Codex cost? GPT-5.3-Codex costs $1.75 per million input tokens and $14 per million output tokens through OpenRouter. In todo.is a typical agent step costs about 6.86 credits ($1 = 100 credits).
- What is the context window of GPT-5.3-Codex? GPT-5.3-Codex can read up to 400K tokens at once and write up to 128K tokens in one reply.
- Can I use GPT-5.3-Codex without coding? Yes. In todo.is you pick GPT-5.3-Codex as your agent's model and write to-dos in plain words, on the web, the Android app, WhatsApp or Telegram. No API key needed: credits are included, or you can bring your own key.
JavaScript is required to use the todo.is app.