GPT 5.2 Codex
GPT 5.2 Codex is served through the Skill Issues marketplace on the OpenAI-compatible and Anthropic-compatible surfaces, with the pricing and limits published below.
gpt-5.2-codex
Text
active
Latency and throughput are withheld until the engine timing semantics are verified against our own request timeline. Sample counts show how much evidence exists for each model.
- Context window
- 272K tokens
- Max output
- 128K tokens
- Surfaces
- anthropic_messages, openai_chat, openai_responses
- Published price
- Unavailable. This model has no published token price yet, so no price is shown rather than an estimate.
- Avg TTFT
- Not measured
- Avg latency
- Not measured
- Avg throughput
- Not measured
- Requests recorded
- 0
Live marketplace prices
Loading live seller prices…
Your requests automatically pick the cheapest eligible seller. Prefer explicit control? See Routing and Combos in your dashboard.
Figures reflect the last 24 hours of measured activity and the currently published price. Compare this model against another on price, context, and measured activity in the model directory, or read the buyer guide before sending a request.