Model Terminal

GPT-5.3-Codex

GPT-5.3-Codex is OpenAI's flagship agentic coding model, designed for long-horizon software engineering tasks including multi-step code generation, debugging, repository work, and tool-using agent workflows. OpenAI describes it as 25% faster than its predecessor and its most capable agentic coding model to date, with benchmark leadership claims on SWE-Bench Pro and Terminal-Bench. Source: supabase.

Value score
Not scored
Context
tokens
Max output
tokens
Price

Capability radar

Peer value bars

Identity

Developer
OpenAI
Openness
Modalities
Release
Knowledge cutoff
Deprecation
API docs

Benchmarks

No benchmark scores yet.

Compare nearby

Pricing

Input / 1M
Output / 1M
Speed