Cost to run
o3-mini
proprietary/o3-mini
- Family
- o3
- Context
- 200,000 tokens
o3-mini can be run self-hosted (rent a GPU + run vLLM/TGI) or through a serverless API (pay per token). Live pricing comparisons:
Self-hosted
Rent a GPU from Lambda, RunPod, Vast.ai, CoreWeave. Full control.
Serverless
Pay per token via Together, Fireworks, DeepInfra, Groq.
Need to benchmark this model against another? Try the calculator or see where it ranks on the InferenceScore leaderboard.