GPT 5.2 Codex is a retired AI model from OpenAI.

GPT 5.2 Codex has a recorded input context length of 400,000 tokens.

The lowest base rates currently recorded across providers for GPT 5.2 Codex are Input Text Tokens at $1.75 per 1M tokens; Output Text Tokens at $14.00 per 1M tokens; and Cache Read Tokens at $0.18 per 1M tokens. The pricing section shows every recorded provider, pricing plan, meter, and condition.

GPT 5.2 Codex is available through the Phaseo API, with 2 active providers currently recorded. The providers section shows the routes and availability currently recorded by Phaseo.

Yes. At least one active provider route for GPT 5.2 Codex currently advertises tool calling support. Provider support can vary, so requests are routed only to compatible routes.

Yes. At least one active provider route for GPT 5.2 Codex currently advertises structured outputs support. Provider support can vary, so requests are routed only to compatible routes.

Phaseo records GPT 5 1 Codex Max as the previous model. View the model family for the complete release history.

Phaseo currently tracks 2 benchmark results for GPT 5.2 Codex. Review the benchmark section for scores, ranks, methodology context, and available sources.

GPT 5.2 Codex accepts image and text inputs. It produces text output.

GPT 5.2 Codex was released on 18 Dec 2025.