OpenAILanguage model

GPT-6 Astra Pro

OpenAI language model.

Common-Sense Reasoning

of 40
1.9 pts behind #1

races led

weeks at #1

86.5

score

The 30-second version

OpenAI language model.

Key takeaways

Latest real results · click any race to watch it

6 Astra ProBest on each axis

Hard Reasoning

#1 frontier6 Astra ProGrok 4.3M3

The distilled version first; open any release for the deeper breakdown.

The GPT Astra family

60-second briefing

6 Astra Pro, explained

Animated summary · captions from the auto-written script

From OpenAI's own guide, distilled

Reasoning models do their own step-by-step thinking; give goals and constraints rather than scripting the chain of thought.
Use the reasoning-effort and verbosity controls instead of prompt tricks to trade depth for speed.
Contradictory instructions waste reasoning tokens; audit system prompts for conflicts.
For agents, define stop conditions and how eager the model should be to keep going without asking.
Structured outputs (JSON schema) beat asking politely for a format.

About OpenAI

Maker of ChatGPT and the GPT series, pursuing general-purpose AI at consumer scale.

Optimising for

Frontier reasoning and math, broad consumer usefulness inside ChatGPT, and tiered models (Astra, Sol, Luna) that trade depth for price and speed.

San Francisco, US · founded 2015

Closest rivals

6 Astra Pro is 3rd in common-sense reasoning; leader is Claude Opus 5.5.