GPT-5.5 Pro
Parameters
-
Context Length
1,000K
Modality
Multimodal
Architecture
Dense
License
Proprietary
Release Date
23 Apr 2026
Knowledge Cutoff
-
Technical Specifications
Attention
Attention Structure
Multi-Head Attention
Attention Heads
-
Key-Value Heads
-
Attention Head Dimension
-
Position Embedding
Absolute Position Embedding
RoPE Theta
-
Sliding Window Attention
-
Sliding Window Size
-
Normalization
-
Activation Function
-
Dimensions
Hidden Dimension Size
-
Number of Layers
-
FFN Intermediate Size (Dense)
-
Multi-Token Prediction Heads
-
Tokenizer
Vocabulary Size
-
GPT-5.5 Pro
GPT-5.5 Pro is OpenAI's maximum-performance variant of GPT-5.5, designed for the hardest questions and highest-accuracy work. Achieves top results on BrowseComp (90.1%), FrontierMath Tier 4 (39.6%), GeneBench (33.2%), and Humanity's Last Exam with tools (57.2%). Particularly strong for business, legal, education, and data science tasks. Available via API as gpt-5.5-pro at $30/M input and $180/M output tokens. Released April 23, 2026.
About GPT-5.5
GPT-5.5 is OpenAI's smartest and most intuitive model yet, and the next step toward a new way of getting work done on a computer. It excels at agentic coding, computer use, knowledge work, and scientific research — areas where progress depends on reasoning across context and taking action over time. It matches GPT-5.4 per-token latency in real-world serving while performing at a significantly higher level of intelligence, and uses fewer tokens to complete the same tasks. State-of-the-art on Terminal-Bench 2.0 (82.7%), SWE-Bench Pro (58.6%), OSWorld-Verified (78.7%), and ARC-AGI-2 (85.0%). Features a 1M context window and costs $5/M input tokens and $30/M output tokens. Released April 23, 2026.
Other GPT-5.5 Models
Evaluation Benchmarks
No evaluation benchmarks for GPT-5.5 Pro available.
Rankings
Overall Rank
-
Coding Rank
-