OpenAI's model naming conventions explained by Ted Sanders
In brief
- Model size is the main driver of performance, cost, and rare fact recall.
- Version numbers combine release order, technical similarities, and improvement scale.
- Minor versions like 6.1 avoid setting expectations as high as major versions like 7.
Ted Sanders from OpenAI explains the logic behind the company's model naming conventions in a blog post. He states that model size remains the primary factor driving performance, reliability, cost per token, and the ability to recall rare facts. Version numbers combine release order, technical similarities, and the scale of improvements, avoiding exaggerated expectations for minor updates like 6.1 compared to major ones like 7.
These conventions aim to clarify model evolution for users without creating disproportionate expectations for minor releases.
This summary gives you the essentials. The full article, its figures and quotes are on Simon Willison's Weblog.
Read the original articleAI-generated summary of the source, checked by rules. How we write · Report an error