ModelPriceWatch.com
Last scan 2026-09-15 Models tracked 259 Providers 32 Cheapest paid Granite 4.0 H Micro $0.017/Mtok in Every price links to its source

GPT-5.2 Codex

by OpenAI

Retired flagship mid tier
Retired on . No longer served on the endpoint we price — the figures below are kept for historical reference only. Replaced by GPT-5.6 Sol. Full retirement status & replacements →
Availability: OpenAI shut this model down on its own API on 2026-07-23 and names GPT-5.6 Sol as the replacement — its deprecations page lists the shutdown under past deprecations. Its model page still quotes $1.75/$0.175/$14 per 1M tokens and no longer carries any deprecation label, and Azure still serves it at that rate — partner platforms set their own retirement schedules, so the price above is what the model last cost, not a rate OpenAI will sell you today.

Today's price · per 1M tokens

Input

$1.75

per 1M tokens

Output

$14.00

per 1M tokens

Blended

$4.81

blended $/1M — 3:1 weighted input:output

Cached input

$0.175

10% of input — prompt caching

Source: official OpenAI pricing · read MODELPRICEWATCH.COM · 2026-09-15

Price receipt

We read OpenAI's own pricing page on and found GPT-5.2-Codex listed with a price on the same row — that read is where the number above comes from. We keep the page text we read; its content hash is 85bce54c0b.

Overview

Retired from the OpenAI API on 2026-07-23; requests to it there now fail. Code-specialized GPT-5.2 variant tuned for agentic software engineering, tool use, and long-horizon coding tasks, with a 400K context, last listed at $1.75/$14 per 1M tokens. Use GPT-5.6 Sol instead.

Capabilities

struck through = not supported
Input 1/5
Text Image Audio Video PDF
Output 1/5
Text Image Audio Video Embedding
Features 3/9
Prompt caching Reasoning Coding Fast inference Long context Open weights Multimodal Web search Realtime

Benchmark performance

accuracy % · higher is better
Percentile vs all tracked models 42.6th2 independent measurements
Percentile per $/Mtok 8.9

We hold no per-benchmark accuracy scores for this model yet, so it has no accuracy average. That is a gap in our coverage, not a sign the model is untested — independent evaluators often publish a composite index for a new model long before releasing its per-benchmark numbers. The percentile above is its standing across the independent composites below.

Independent composite scores
  • AA Intelligence Index: 40.1 (Artificial Analysis composite across reasoning, knowledge and coding evals)
  • LMSYS Chatbot Arena: 1331 ELO (Human preference)
How it stacks up
  • Ranks #93 of 181 comparably-measured models by percentile score, across 2 independent measurements
  • Ranks #155 of 181 comparably-tested models by normalized performance per dollar

123 tokens/sec output

Source: Artificial Analysis (via BenchLM), LMSYS Chatbot Arena (snapshot via BenchLM) · updated · See full rankings →

Specifications

Provider
OpenAI
Context window
400K tokens
Modality
text
Parameters
Proprietary
Open source
No — proprietary
Released
Status
Retired ended
Last updated
Tags
codingflagshipretired

Availability verified: per OpenAI's own deprecation notice