OpenAI API Implements 40% Price Increase for GPT-5.2 Output and Input Tokens, Renames GPT-5 to GPT-5.2

Between the Oct 4, 2025 and Jan 5, 2026 captures of OpenAI's pricing page, Pulse recorded 11 changes: AgentKit platform added - Build, deploy, and optimize production-grade agents with Agent Builder, ChatKit, and Evals. 1GB free tier for storage, $0.10/GB-day beyond free tier, billing begins November 1, 2025; GPT-5.2 output price increased from $10.000/1M tokens to $14.000/1M tokens (40% increase); GPT-5 nano removed - The fastest, cheapest version priced at $0.050/1M input tokens and $0.400/1M output tokens; GPT-image-1-mini model added for image generation at $2.00/1M input tokens (text), $2.50/1M input tokens (image), and $8.00/1M output tokens (image); Sora Video API added - Richly detailed, dynamic video generation with models sora-2 and sora-2-pro, pricing from $0.10 to $0.50 per second based on resolution; GPT-image-1.5 model added for image generation with text input at $5.00/1M input tokens, $10.00/1M output tokens and image input at $8.00/1M input tokens, $32.00/1M output tokens; GPT-5.2 input price increased from $1.250/1M tokens to $1.750/1M tokens (40% increase); GPT-5 renamed to GPT-5.2; gpt-realtime-mini text cached input price decreased from $0.30/1M tokens to $0.06/1M tokens (80% decrease); GPT-5.2 cached input price increased from $0.125/1M tokens to $0.175/1M tokens (40% increase); GPT-5.2 pro plan added - The smartest and most precise model at $21.00/1M input tokens and $168.00/1M output tokens. Pulse files it as a pricing change (feature added, price increase, plan removed, plan renamed, price decrease, new plan) with a neutral direction. Pulse's latest capture (Nov 9, 2026) lists gpt-image-2.5-flare at $2 per 1M cached image input tokens + $30 per 1M image output tokens, gpt-6.1-sol at $0.20 per 1M cached input tokens (long context) + $2 per 1M input tokens (short context), gpt-6-luna at $0.25 per 1M cache write tokens (long context) + $0.20 per 1M input tokens (long context), gpt-5.6-sol at $30 per 1M output tokens (long context) + $8 per 1M input tokens (long context) and gpt-6-astra at $20 per 1M input tokens (long context) + $1 per 1M cached input tokens (short context).

OpenAI's previous recorded change was "Openai API Introduces GPT-5 Mini and Nano Models with Significantly Lower Pricing, Enhancing Accessibility for Users" (Q3 2025). OpenAI has been tracked since 2023-11-14.

In the same window Metabase and Anthropic also changed their AI & Machine Learning pricing pages. Those moves are listed further down, and the Pulse Agent can compare them side by side.