O1 Mini: benchmark results
OpenAI's smaller o1-series reasoning model for cost-efficient STEM tasks. Provider: OpenAI. Released 2024-09-12. Access: API.
Aggregate Rating 1643 ± 7, rank #323 of 1656 rated models, from 205 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AgentIF | 59.8 | Constraint Success Rate (imported) | 100 |
| Carlini Applied LLM Benchmark | 62 | Pass Rate (%) | 100 |
| EnterpriseBench Sandbox (DSPy, CoT) | 0.33 | Judged task accuracy (0-1; 500 enterprise tasks, final execu | 100 |
| EnterpriseBench Sandbox (DSPy, Gold Plan) | 0.62 | Judged task accuracy (0-1; 500 enterprise tasks, final execu | 100 |
| EnterpriseBench Sandbox (DSPy, No Planning) | 0.29 | Judged task accuracy (0-1; 500 enterprise tasks, final execu | 100 |
| EnterpriseBench Sandbox (DSPy, ReAct) | 0.38 | Judged task accuracy (0-1; 500 enterprise tasks, final execu | 100 |
| EnterpriseBench Sandbox (LangChain, CoT) | 0.28 | Judged task accuracy (0-1; 500 enterprise tasks, final execu | 100 |
| EnterpriseBench Sandbox (LangChain, Gold Plan) | 0.51 | Judged task accuracy (0-1; 500 enterprise tasks, final execu | 100 |
| EnterpriseBench Sandbox (LangChain, ReAct) | 0.35 | Judged task accuracy (0-1; 500 enterprise tasks, final execu | 100 |
| Omni-MATH | 60.54 | Overall Accuracy (%) | 100 |
| ProLLM - Q&A Assistant | 98.9 | Score (%) | 100 |
| SEAL - Coding | 1237 | Score | 100 |
Interactive version: theaggregate.ai/model?slug=o1-mini · How It Works · Data refreshed daily, snapshot 2026-10-20.