OpenAI
GPT-5.3 Codex (xhigh)
GPT-5.3 Codex (xhigh) is an OpenAI GPT-5 family model, positioned around stronger reasoning, coding, factuality, and agentic work than earlier GPT generations. This profile keeps the release context separate from the local Artificial Analysis measurements so you can compare the specific reasoning tier on benchmark quality, speed, and cost.
Introducing GPT-5Operational Metrics
Output Speed87.1 tok/s
First Token69.23s
Blended Price$4.81/M
Model Metadata
Queryable facts extracted from the upstream model payload.
ReleaseFeb 5, 2026
Context Windown/a
Modalitiesn/a
Strength: GPQA
Rank #6 across 478 models.
91.5%
Strength: HLE
Rank #6 across 474 models.
39.9%
Strength: Coding
Rank #7 across 410 models.
53.1
Watch Area: TTFT
Rank #287 across 293 models.
69.23s
Watch Area: Output $
Rank #283 across 325 models.
$14.00/M
Watch Area: Blended $
Rank #281 across 325 models.
$4.81/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 | 53.6 | #9 |
| Artificial Analysis Coding Index | coding | 53.1 | #7 |
| GPQA | reasoning | 91.5% | #6 |
| Humanity's Last Exam | reasoning | 39.9% | #6 |
| SciCode |