OpenAI
GPT-5.4 Pro
A model focused on pure reasoning and mathematical problem-solving, setting a record on the FrontierMath benchmark.
Our editorial assessment as of Jun 4, 2026, based on the public sources listed below.
Overall rank
#9
Tier
S
Score
88.0
Main category
Reasoning
Strengths
- Leads in mathematical problem-solving capabilities.
- Benefits from the largest ecosystem of plugins, integrations, and third-party tools.
- High scores on reasoning and coding benchmarks like SWE-bench.
Considerations
- The most expensive model in some rankings.
- Lacks a clear overall performance lead over newer, cheaper competitors.
Price
API $5 / 1M tokens (input), $25 / 1M tokens (output)
Score breakdown
Performance
96.0
Reputation
88.0
Price
40.0
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
- Introducing GPT-5.4 - OpenAIAccessed: 2026-06-03
- Introducing GPT-5.5 - OpenAI (GPT-5.4 Pro benchmark table)Accessed: 2026-06-03
- AiZolo Top AI Models March 2026Accessed: 2026-04-20
- OpenAI Model Release NotesAccessed: 2026-04-20