/ Models Index / OpenAI o3

OpenAI o3

Reasoning · OpenAI · OpenAI · Released Jan 2025
A

Excellent (A) — Capability Grade

OpenAI o3 is the high-compute reasoning model. State-of-the-art on math (AIME, FrontierMath) and competitive programming. The original demonstration of the test-time-compute scaling paradigm. Per-token inference cost is materially higher than non-reasoning models; reserve for genuine reasoning-required workloads.

91
Composite / 100
/ Subscore Breakdown · 6 Capability Dimensions

Where this grade comes from.

General Reasoning
A
Code Generation
A
Math & STEM
A
Tool Use & Agency
B+
Multimodal
B-
Safety & Alignment
B+
/ Key Events & Disclosures

Release timeline & positioning.

  • Released Jan 2025
  • Test-time-compute scaling demonstration
  • State-of-art on FrontierMath
  • High inference cost

/ Best for

Math, scientific reasoning, and competitive programming tasks where extra inference compute justifies the cost.

/ Watch out for

Latency and per-token cost materially higher than non-reasoning models. Not suitable for high-throughput production where standard models suffice.