OpenAI
GPT-5.5
OpenAI's April 2026 frontier release leading internal SWE and agentic coding evaluations.
Our editorial assessment as of Jun 4, 2026, based on the public sources listed below.
Overall rank
#3
Tier
S
Score
92.0
Main category
Coding
Strengths
- Tops Terminal-Bench 2.0 and GDPval among reported peers
- Strong conceptual grasp of large codebases per early testers
- Improved inference efficiency over GPT-5.4
Considerations
- Higher API cost than mid-tier variants
- Some reports cite elevated hallucination rates on edge prompts
- Very recent release; long-term reliability still being assessed
Price
ChatGPT Pro tier · API ~$11/1M output tokens
Score breakdown
Performance
96.0
Reputation
92.0
Price
40.0
Recency
100.0
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
- Model Release Notes | OpenAI Help CenterAccessed: 2026-06-03
- Introducing GPT-5.5 - OpenAIAccessed: 2026-06-03
- Introducing GPT-5.5 | OpenAIAccessed: 2026-05-13