O-series reasoning model for hard analysis, math, coding, and planning
Specification
most-agreed values
Context
128K
Max output
66K
Released
2025-01-01
Knowledge cutoff
2025-01
Retires
—
Open weights
no
Input
text
Output
text
Price
US dollars per million tokens · most-agreed
Input
$1.10
Output
$4.40
Cache read
$0.55
Batch and priority prices
US dollars per million tokens · per listing
| Price | Listing | $/M |
|---|---|---|
| Batch input | azure | $0.605 |
| Batch output | azure | $2.42 |
Batch input is 50% below the interactive input price at the same listing.
Quality
5 benchmarks · 2 sources
Composite
49th
percentile of 334 scored models
Rank
172 / 334
±0.05 sd
Evidence
5 × 2
3 or more benchmarks, from 2 or more sources
Effort range
—
one configuration only
coding26th1/3 bench
reasoning53rd1/4 bench
math55th3/7 bench
Every score
one row per source and configuration — nothing averaged away
| Benchmark | Score | Configuration | Source | Run |
|---|---|---|---|---|
| MATH level 5 | 86.7 ±0.7 | default · 2 runs | Epoch AI | 2025-02-13 |
| GPQA diamond | 60.9 ±2.7 | default · 2 runs | Epoch AI | 2025-02-13 |
| OTIS Mock AIME 2024-2025 | 45.8 ±6.5 | default · 2 runs | Epoch AI | 2025-03-06 |
| Aider polyglot | 32.9 | default · aider whole | Aider | 2024-12-22 |
| FrontierMath-v1 | 0.8 ±0.0 | default · 4 runs | Epoch AI | 2025-03-06 |
The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.
Available from 3 hosts
| Host | In $/M | Out $/M | Cache rd | Cache wr | Context | Output | Retires |
|---|---|---|---|---|---|---|---|
| helicone | $1.10 | $4.40 | $0.55 | — | 128K | 66K | — |
| replicate · replicate | $1.10 | $4.40 | — | — | — | — | — |
| azure | $1.21 $1.10 | $4.84 $4.40 | $0.605 $0.55 | — | 128K | 66K | — |
More from OpenAI
most-hosted first