OpenAI
o3
o3 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 Speed72.7 tok/s
First Token8.55s
Blended Price$3.50/M
Model Metadata
Queryable facts extracted from the upstream model payload.
ReleaseApr 16, 2025
Context Windown/a
Modalitiesn/a
Strength: MATH-500
Rank #3 across 201 models.
99.2%
Strength: AIME
Rank #8 across 194 models.
90.3%
Strength: LCB
Rank #26 across 343 models.
80.8%
Watch Area: TTFT
Rank #254 across 293 models.
8.55s
Watch Area: Input $
Rank #275 across 325 models.
$2.00/M
Watch Area: Blended $
Rank #268 across 325 models.
$3.50/M
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 | 38.4 | #87 |
| Artificial Analysis Coding Index | coding | 38.4 | #57 |
| Artificial Analysis Math Index | math | 88.3 | #35 |
| MMLU-Pro | reasoning | 85.3% | #29 |