Overview
A massive multimodal model.
Key Capabilities
●
Historic performance
●
Multimodal
●
Reasoning jump
Benchmarks
MMLU Score
86.4%
HumanEval (Coding)
67%
MATH Benchmark
52%
Learn to Master This Model
Take our free structured GPT-4 course — from basics to advanced techniques.
Technical Specs
Context8,192 tokens
Paramsunknown
LicenseProprietary
ArchTransformer
Prompt Library
Browse Coding Prompts →
Previous Version
Gpt 3 5 →