OpenAI
GPT-5.3 Codex
An agentic coding model that combines code generation with general-purpose intelligence.
Our editorial assessment as of Jun 4, 2026, based on the public sources listed below.
Overall rank
#14
Tier
S
Score
87.0
Main category
Coding
Strengths
- Described as a 'self-improving' agentic coding model.
- Sets new highs on key coding benchmarks.
- Approximately 25% faster than previous versions.
Considerations
- Specialized for coding, may not be the top choice for general reasoning.
- Multiple versions of GPT-5 can be confusing for users.
Price
API $1.75 / 1M tokens (input), $14.00 / 1M tokens (output)
Score breakdown
Performance
92.0
Reputation
88.0
Price
60.0
Recency
95.0
Verification status
Each field carries its own re-verification date. Amber means older than 90 days — treat those figures as stale until re-checked.
Sources
- OpenAI Model Release NotesAccessed: 2026-06-03
- Introducing upgrades to Codex - OpenAIAccessed: 2026-06-03
- Frontier AI Models 2026: GPT-5.3, Claude 4.6, G - TeamDay.aiAccessed: 2026-05-26
- AI Benchmarks 2026: Top Evaluations and Their LimitsAccessed: 2026-05-13
- AI Leaderboard 2026Accessed: 2026-04-20
- OpenAI — ChatGPT Release NotesAccessed: 2026-04-19
- OpenAI API PricingAccessed: 2026-04-19