O3 MINI
OpenAI o3-mini is a cost-efficient language model optimized for STEM reasoning tasks, particularly excelling in science, mathematics, and coding. This model supports the `reasoning_effort` parameter, which can be set to "high", "medium", or "low" to control the thinking time of the model.
Input
Output
Context
200K
Max Output
100K
Parameters
—
Technical Specifications
Model TypeChat
Context Window200K
Max Output Tokens100K
Parameters—
Training Cutoff—
Licenseproprietary
Capabilities
Input Modalities
textfile
Output Modalities
text
Features
chain_of_thoughtcode_generationextended_thinkingfunction_callingreasoning
Benchmarks
Chatbot Arena
#70
Rank (Text)
+3/-3
arena text ci 95
1318
Arena Score (Text)
43,216
Votes (Text)
Artificial Analysis
8.7%
HLE
77.0%
AIME
74.8%
GPQA
39.9%
SciCode
97.3%
MATH 500
79.1%
MMLU Pro
17.9
Coding
71.7%
LiveCodeBench
25.9
Intelligence
156
Speed (tok/s)
5.9
TTFA (s)
5.9
TTFT (s)
Resources & Links
proprietary — Proprietary - API access only
Related Tools
Data sourced from official provider APIs and documentation
Last updated: May 5, 2026
Start building with the right model.
From model selection to production, one platform, no fragmentation.