modelgrep
O

OpenAI: GPT-5.3-Codex

openai/gpt-5.3-codex

9th smartest of 178ReasoningToolsJSONVision
Use via OpenRouter ↗
Intelligence
53.6
9th of 178
Design Elo
1229
Game Dev
Speed
53
158th fastest
Latency
4.2s
first token
Input price
$1.75
235th cheapest
Context
400K
128K max out

How it compares

Smarter than95%
of all ranked models
Faster than46%
of all ranked models
Cheaper than21%
of all ranked models

Overview

GPT-5.3-Codex is OpenAI’s most advanced agentic coding model, combining the frontier software engineering performance of GPT-5.2-Codex with the broader reasoning and professional knowledge capabilities of GPT-5.2. It achieves state-of-the-art results...

Benchmarks

independent · via OpenRouter
Artificial Analysis95th percentile
Intelligence Index
53.6
Coding Index
53.1
Agentic Index
60.5
GPQA Diamond
92%
Humanity's Last Exam
40%
SciCode
53%
Tau²-Bench (agentic)
86%
Design Arena · Elo11,845 tournaments
Game Dev
1229
Website
1208
Data Viz
1207
codecategories
1196
UI Component
1194
asciiart
1193
svg
1186
godotgamedev
1162
mobileapps
1142
webapps
1132
3D
1089
androidnative
1084
fullstack
1062

Providers & pricing (2)

ProviderIn $/MOut $/MUptime
OpenAI$1.75$14.0099.9%
Azure$1.75$14.00100%

Specifications

Context window400K
Max output128K
Knowledge cutoff
Input modalitiestext, image, file
Output modalitiestext
Prompt caching
Cache read price$0.175/M
ModeratedYes

GPT-5.3-Codex FAQ

How much does GPT-5.3-Codex cost?

GPT-5.3-Codex costs $1.75 per million input tokens and $14.00 per million output tokens via OpenRouter, making it 235th cheapest of 298 paid models.

How smart is GPT-5.3-Codex?

GPT-5.3-Codex scores 53.6 on the Artificial Analysis Intelligence Index, ranking 9th of 178 benchmarked models, with a GPQA Diamond score of 92%.

How fast is GPT-5.3-Codex?

GPT-5.3-Codex generates around 53 tokens per second with 4.2s time-to-first-token (p50), the 158th fastest tracked model.

What is GPT-5.3-Codex's context window?

GPT-5.3-Codex supports a 400K-token context window and can output up to 128K tokens. It accepts text, image, file input.

Compare head-to-head