Reasoning models
Models that think before they answer. Single steps get much smarter.
stageBuilt today Open on the map →
Facts
- Lane
- Models and thresholds
- Wave
- 5 of 12 (build order, not a date)
- Needs
- 10 boxes all the way back through hard links
- Unlocks
- 60 boxes downstream through hard links
- Leverage
- None: already built
- Cluster
- Science and medicine
- When
- Here since 2024
Evidence as of October 2026
Level 2, "Reasoners", on the five-level scale toward AGI that OpenAI reportedly uses internally.
Watch
Price and speed of frontier reasoning at a fixed benchmark score.
Timeline note
OpenAI’s o1 reasoning model, September 2024.