← All models
openai
OpenAI: o3 Pro
The o-series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o3-pro model uses more compute to think harder and provide consistently better answers. Note that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations
200,000 context
Modalities:text, image, file->text
Released:6/10/2025
The o-series of models are trained with reinforcement learning to think before they answer and perform complex reasoning. The o3-pro model uses more compute to think harder and provide consistently better answers. Note that BYOK is required for this model. Set up here: https://openrouter.ai/settings/integrations
Weekly tokens
5.3M
Tokens generated this week (network-wide)
Rankings (last periods)
No ranking data yet for this model.