OpenAI
o3-mini
o3-mini is one of OpenAI's reasoning-focused models, built for harder multi-step tasks where deliberate problem solving matters more than simple chat completion. The benchmark snapshot highlights how that reasoning emphasis translates into scores, latency, and value versus general-purpose models.
Introducing o3 and o4-miniOperational Metrics
Output Speed140.1 tok/s
First Token9.18s
Blended Price$1.93/M
Model Metadata
Queryable facts extracted from the upstream model payload.
ReleaseJan 31, 2025
Context Windown/a
Modalitiesn/a
Strength: MATH-500
Rank #24 across 201 models.
97.3%
Strength: AIME
Rank #27 across 194 models.
77.0%
Strength: LCB
Rank #61 across 343 models.
71.7%
Watch Area: TTFT
Rank #258 across 293 models.
9.18s
Watch Area: Input $
Rank #221 across 325 models.
$1.10/M
Watch Area: Value
Rank #218 across 323 models.
13.5
Strength Profile
Percentile score by analysis domain.
* Cost is inverted: lower input, output, and blended prices rank higher.
Benchmark Percentiles
Higher bars mean stronger relative placement.
All Benchmarks
| Metric | Domain | Value | Rank |
|---|---|---|---|
| Artificial Analysis Intelligence Index | overall | 25.9 | #180 |
| Artificial Analysis Coding Index | coding | 17.9 | #211 |
| MMLU-Pro | reasoning | 79.1% | #129 |
| GPQA | reasoning | 74.8% | #161 |
| Humanity's Last Exam |