Checked against the OpenRouter public models feed. Prices are copied, not estimated.

The Best Model ForWhich AI model to use, from ten labs.

GPT-5.2-Codex vs GPT-6.2 Astra

Short answer

GPT-5.2-Codex ($1.75 in / $14.00 out); GPT-6.2 Astra ($8.00 in / $40.00 out). Lowest listed input price: GPT-5.2-Codex. Longest listed context: GPT-6.2 Astra at 1,050,000 tokens. A lower price is not a quality ranking.

Side-by-side facts. The third-party index is copied when the feed has one. Unknown means it did not.
FactGPT-5.2-CodexGPT-6.2 Astra
LabOpenAIOpenAI
KindChatChat
Listed14 Jan 202617 Nov 2026
Context400,0001,050,000
Input / million$1.75$8.00
Output / million$14.00$40.00
Batch inputNot listed$4.00
Batch outputNot listed$20.00
Inputsimage, textfile, image, text
Outputstexttext
ToolsYesYes
Open weightsNo public weights id listedNo public weights id listed
Knowledge cutoffUnknownUnknown
Third-party indexUnknown62.1
GuideNew, not yet reviewedReviewed

Input and output prices are USD per million tokens, from pricing.prompt and pricing.completion. The index is not our benchmark.

GPT-5.2-Codex

New, not yet reviewed. No guide note yet.

GPT-5.2-Codex is listed at $1.75 input and $14.00 output per million tokens, with a 400,000-token context window.

GPT-6.2 Astra

OpenRouter describes GPT-6.2 Astra as OpenAI's flagship for the hardest projects, including engineering and scientific research. Use it when a wrong answer is expensive to redo. The feed lists Astra Pro as the same model with a higher reasoning mode, and warns that pro mode spends more, even though the published per-token rates match.

GPT-6.2 Astra is listed at $8.00 input and $40.00 output per million tokens, with a 1,050,000-token context window.

Choose models

Popular comparisons