O-series reasoning model for hard analysis, math, coding, and planning
Specification
openai listing
Context
200K
Max output
100K
Released
2024-12-05
Knowledge cutoff
2023-09
Retires
2026-10-23
Open weights
no
Input
text, image, pdf
Output
text
Price
US dollars per million tokens · openai list price
Input
$15.00
Output
$60.00
Cache read
$7.50
Cache write
free
Batch and priority prices
US dollars per million tokens · per listing
| Price | Listing | $/M |
|---|---|---|
| Batch input | azure | $8.25 |
| Batch input | azure | $7.50 |
| Batch output | azure | $33.00 |
| Batch output | azure | $30.00 |
Batch input is 45% below the interactive input price at the same listing.
Quality
9 benchmarks · 3 sources
Composite
52nd
percentile of 344 scored models
Rank
165 / 344
±0.04 sd
Evidence
9 × 3
3 or more benchmarks, from 2 or more sources
Effort range
—
one configuration only
coding63rd2/3 bench
reasoning42nd2/4 bench
math45th4/7 bench
knowledge46th1/1 bench
Every score
one row per source and configuration — nothing averaged away
| Benchmark | Score | Configuration | Source | Run |
|---|---|---|---|---|
| MATH level 5 | 90.3 ±0.6 | default · 3 runs | Epoch AI | 2025-02-13 |
| Aider polyglot | 61.7 | high · aider diff | Aider | 2024-12-21 |
| SWE-Bench verified | 64.6 | default · W&B Programmer O1 crosscheck5 | SWE-bench | — |
| GPQA diamond | 69.3 ±3.1 | default · 4 runs | Epoch AI | 2026-07-20 |
| OTIS Mock AIME 2024-2025 | 57.8 ±7.5 | default · 4 runs | Epoch AI | 2026-07-15 |
| SimpleQA Verified | 41.1 ±1.6 | default | Epoch AI | 2026-08-31 |
| FrontierMath-v1 | 4.7 ±1.7 | default · 2 runs | Epoch AI | 2025-03-07 |
| Chess Puzzles | 9.0 ±3.0 | default · 3 runs | Epoch AI | 2026-10-13 |
| FrontierMath-Tiers-1-3-v2 | 11.1 ±2.1 | default · 3 runs | Epoch AI | 2026-08-28 |
The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.
Available from 11 hosts
| Host | In $/M | Out $/M | Cache rd | Cache wr | Context | Output | Retires |
|---|---|---|---|---|---|---|---|
| openaivendor | $15.00 | $60.00 | $7.50 | — | 200K | 100K | 2026-10-23 |
| poe | $14.00 | $54.00 | — | — | 200K | 100K | — |
| azure | $15.00 $16.50 | $60.00 $66.00 | $7.50 $8.25 | — | 200K 128K | 100K 33K | 2026-11-19 2025-07-28 |
| azure-cognitive-services | $15.00 | $60.00 | $7.50 | — | 200K | 100K | — |
| helicone | $15.00 | $60.00 | $7.50 | — | 200K | 100K | — |
| impossibl | $15.00 | $60.00 | $7.50 | — | 200K | 100K | — |
| merge-gateway | $15.00 | $60.00 | $7.50 | — | 200K | 100K | — |
| nano-gpt | $15.00 | $60.00 | $7.50 | — | 200K | 100K | — |
| replicate · replicate | $15.00 | $60.00 | — | — | — | — | — |
| vercel | $15.00 | $60.00 | $7.50 | — | 200K | 100K | — |
| vercel_ai_gateway · vercel-ai-gateway | $15.00 | $60.00 | $7.50 | free | 200K | 100K | — |
Price history
append-only observations · a listing writes a row only when its price moves
| Date | Host | Field | $/M |
|---|---|---|---|
| 2026-09-02 | openrouter | Cache read | $7.50 |
| 2026-10-10 | openrouter | Cache read | $7.50 |
| 2026-09-02 | openrouter | Input | $15.00 |
| 2026-10-10 | openrouter | Input | $15.00 |
| 2026-09-02 | openrouter | Output | $60.00 |
| 2026-10-10 | openrouter | Output | $60.00 |
| 2026-10-13 | aimlapi | Cache read | $9.75 |
| 2026-10-23 | aimlapi | Cache read | $7.50 |
| 2026-10-13 | aimlapi | Input | $19.50 |
| 2026-10-23 | aimlapi | Input | $15.00 |
| 2026-10-13 | aimlapi | Output | $78.00 |
| 2026-10-23 | aimlapi | Output | $60.00 |
More from OpenAI
most-hosted first