GPT-5.2-Codex

OpenAI·Degraded·Released 2025-12-18textvision
Context400k
Max output128k
Input / 1M$1.75 ↗
Output / 1M$14 ↗
Cached in$0.175
Benchmarks13

Benchmarks

Cited public sources · rank is exact-benchmark, apples-to-apples

LiveBench Agentic Coding (LiveBench)#35/5649.4% ↗
LiveBench Coding (LiveBench)#6/5683.6% ↗
LiveBench Data Analysis (LiveBench)#20/5678.2% ↗
LiveBench Instruction Following (LiveBench)#36/5666.4% ↗
LiveBench Language (LiveBench)#47/5673.7% ↗
LiveBench Mathematics (LiveBench)#33/5688.8% ↗
LiveBench Reasoning (LiveBench)#47/5677.7% ↗
Cybersecurity Capture The Flag Challenges#2/267.4% ↗
OSWorld-Verified#19/2038.2% ↗
SWE-Lancer IC Diamond#2/276% ↗
SWE-Bench Pro#13/2156.4% ↗
Terminal-Bench 2.0#5/2064% ↗

GPT-5.2-Codex FAQ

What is GPT-5.2-Codex?
GPT-5.2-Codex is a large language model from OpenAI, released on December 18, 2025. It accepts text and image input.
When was GPT-5.2-Codex released?
OpenAI released GPT-5.2-Codex on December 18, 2025.
How much does GPT-5.2-Codex cost?
GPT-5.2-Codex costs $1.75 per 1M input tokens and $14 per 1M output tokens through the OpenAI API. Cached input costs $0.175 per 1M tokens. Prices are from OpenAI's official pricing as of October 5, 2026.
What is GPT-5.2-Codex's context window?
GPT-5.2-Codex has a 400K-token context window and can generate up to 128K output tokens.
How does GPT-5.2-Codex score on benchmarks?
GPT-5.2-Codex has 13 cited benchmark scores, including LiveBench Agentic Coding (LiveBench) 49.4%, LiveBench Coding (LiveBench) 83.6% and LiveBench Data Analysis (LiveBench) 78.2%. Each score links to its source.

Access GPT-5.2-Codex and every other model through one endpoint with automatic failover: Respan gateway.